journal of financial transformation
Payments Emerging models Transition Future
Recipient of the APEX Awards for Publication Excellence 2002-2004
12/2004/#12
journal the
Our 200 mph laboratory.
bmw williamsf1 team
Š2003 Hewlett-Packard Development Company, L.P.
The BMW WilliamsF1 Team chose HP to provide the supercomputer used to design the car and to conduct thousands of race simulations. And before the car even hits the track, HP servers and notebooks are used to analyze research data that enables the team to make precise suspension and engine adjustments. It’s mission-critical computing for fast-moving enterprises, and then some. www.hp.com/plus_bmwwilliamsf1
Editor Shahin Shojai, Director of Strategic Research, Capco Advisory Editors Predrag Dizdarevic, Partner, Capco Bill Irving, President, Capco John Owen, Partner, Capco Editorial Board Franklin Allen, Nippon Life Professor of Finance, The Wharton School, University of Pennsylvania Joe Anastasio, CEO, Cross Border Exchange, and Partner, Capco Philippe d’Arvisenet, Group Chief Economist, BNP Paribas Jacques Attali, Chairman, PlaNet Finance Rudi Bogni, Former Chief Executive Officer, UBS Private Banking Bruno Bonati, Member of the Executive Board and Division Head Technology & Operations, Credit Suisse Financial Services David Clark, NED on the board of financial institutions and a former senior advisor to the FSA Géry Daeninck, former CEO, Robeco Douglas W. Diamond, Merton H. Miller Distinguished Service Professor of Finance, Graduate School of Business, University of Chicago Elroy Dimson, Professor of Finance, London Business School Nicholas Economides, Professor of Economics, Leonard N. Stern School of Business, New York University Michael Enthoven, Chief Executive Officer, NIB Capital Bank N.V. José Luis Escrivá, Group Chief Economist, Grupo BBVA George Feiger, Executive Vice President and Head of Wealth Management, Zions Bancorporation Gregorio de Felice, Group Chief Economist, Banca Intesa Wilfried Hauck, Chief Executive Officer, Allianz Dresdner Asset Management International GmbH Thomas Kloet, Senior Executive Vice-President & Chief Operating Officer, Fimat USA, Inc. Herwig Langohr, Professor of Finance and Banking, INSEAD Mitchel Lenson, Global Head of Operations & Technology, Deutsche Bank Group David Lester, Chief Information Officer, The London Stock Exchange Donald A. Marchand, Professor of Strategy and Information Management, IMD and Chairman and President of enterpriseIQ® Colin Mayer, Peter Moores Professor of Management Studies, Saïd Business School, Oxford University Robert J. McGrail, Chairman of the Board, Omgeo Jeremy Peat, Group Chief Economist, The Royal Bank of Scotland Jos Schmitt, Partner, Capco Kate Sullivan, Chief Operating Officer, e-Citi John Taysom, Founder & Joint CEO, The Reuters Greenhouse Fund Graham Vickery, Head of Information Economy Unit, OECD Norbert Walter, Group Chief Economist, Deutsche Bank Group David Weymouth, Chief Information Officer, Barclays Plc
Table of contents Nobel Laureate View 6 Nash’s theories and their impact on economics and finance A discussion with Prof. John F. Nash Jr., Senior Research Mathematician, Department of Mathematics, Princeton University, and Joint-Winner of The Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel 1994
Emerging models 12 Opinion: Bringing broadband banking to Iraq
85
16 Opinion: Unifying the European payments landscape Philippe Menier, Chief Operating Officer, Visa Europe Ltd
21 Opinion: Credit card security on the Net: Where is it today?
93
24 Opinion: Broadening the value proposition around payments outsourcing 26 Opinion: Bringing payments together Michel Akkermans, Chairman and CEO, Clear2Pay
31
Future 104 Opinion: Innovation on networks: Coordination, governance, and the case of VISA Matthew Cardillo, Assistant Economist, Federal Reserve Bank of Kansas City Antoine Martin, Economist, Federal Reserve Bank of Kansas City Michael J. Orlando, Senior Economist, Federal Reserve Bank of Kansas City
Payment systems and regulation Hans Geiger, Professor, The Swiss Banking Institute, University of Zurich Fritz Klein, Independent Consultant, and Chairman, CLS Holdings/ CLS Bank International
39 Why has stored-value not caught on? Sujit Chakravorti, Senior Economist, Federal Reserve Bank of Chicago
49 Online payments systems for e-commerce Caroline Paunov, Consultant, OECD Graham Vickery, Head of the Information Economy Unit, OECD
108 Opinion: The evolution of currency James Turk, founder, GoldMoney.com
110 Opinion: Automating payment processes to reduce working capital Kurt Cavano, Chairman and CEO, TradeCard
112 Opinion: New agenda for payment service providers Terry Dirienzo, Director, Group Payments & Settlements, Barclays Plc. Mark Hale, Head of Payments Strategy, Group Payments & Settlements, Barclays Plc.
55 Six smart moves when playing the smart card game Leo Van Hove, Assistant Professor of Economics, Vrije Universiteit Brussel (Free University of Brussels)
Transition 64 Opinion: Payments policy in the information age Ronald J. Mann, Ben H. & Kitty King Powell Chair in Law, Co-Director, Center for Law, Business & Economics, The University of Texas School of Law
68 Opinion: Creating a profitable infrastructure — The payments challenge for banks
116 Opinion: New models of collaboration in transaction banking Wolfgang Gaertner, Chief Information Officer, Global Banking Division - Global Transaction Banking, Deutsche Bank, and, Director, SWIFT
120 Opinion: European payment service providers: A race to market and the nature of future victory? Julian Wakeham, Partner, Capco Andrew Hogan, Managing Principal, Capco Hector Nelson, Consultant, Capco
Ann Cairns, Global Head, Working Capital, ABN AMRO
71 Opinion: Payments in transition: Where have all the changes gone?
125 Network-based payments and e-settlement Harry Leinonen, Adviser to the Board, Bank of Finland
Mark Webster, Partner, Capco
74 Opinion: Electronic payments: The drive towards a competitive, customer service driven utility Rod Dew, Marketing Director, Distra Pty Ltd
79 Simulation: A powerful research tool in payment and settlement systems Harry Leinonen, Adviser to the Board, Bank of Finland Kimmo Soramäki, Policy Expert, European Central Bank
Technological innovation in retail payments: Key developments and implications for banks Karen Furst, Policy Analyst, Office of the Comptroller of the Currency Daniel E. Nolle, Senior Financial Economist, Office of the Comptroller of the Currency
Sankarson Banerjee, Senior Architect, Mphasis
Thomas Halpin, Vice President, JPMorgan Treasury Services
European payment systems and monetary union Francisco J. Callado Muñoz, Assistant Professor, University of Girona Natalia Utrero González, Visiting Professor, Universitat Autonoma de Barcelona
Kathleen Tyson-Quah, Chief Executive, Granularity Ltd
131
Banks’ strategies for payment services: Which role for debit cards? Francesco Saita, Associate Professor, Financial Markets and Institutions Department, Università Bocconi
141 Delivering migrant workers’ remittances Roger Ballard, Director, Centre for Applied South Asian Studies, University of Manchester
155 Check 21 and the migration to electronic payments Adam Dener, Partner, Capco
Payments finally hits the boardroom table
Overlooked and ignored for years — if not decades — at a time, the payments industry has been in a constant state of flux as of late. Payments services providers, such as universal banks, have had to cope with the notion that a business possibly generating handsome returns has now more often than not become a cost center. In my opinion there is, therefore, no better moment to dedicate an entire issue of the Journal to payments in all its facets. As with all other transitions in the financial services sector, change in the payments industry is a global phenomenon with local or regional flavoring: In Europe the introduction of the single currency as legal tender in 2002 was immediately followed by a decree forcing payments service providers to equalize the cost borne by users of making a cross-border euro payment to the cost of making a domestic euro payment: zero. The custom of paper-check payment in the United States — a business estimated to generate U.S$25 billion in customer support account fees and U.S.$34 billion in deposit/float fees — is on a continued course of change to electronic transaction processing, enhanced by the Check Clearing for the 21st Century Act. Asia’s various developed and established economies see an increased presence of credit and debit card providers as an alternative to cash payments, and ‘secondary’ payments systems used by migrant workers to send funds back home. These developments help dramatically change payments’ traditional low visibility and strategic importance in banks, and ensure that the cost implications of maintaining a payments capability in-house and the value of the annuity revenue they develop become increasingly recognized as important. Concepts which since many years have dominated the agenda of securities trading and processing providers — shared service centers, outsourcing and off-shoring, to name but a few — are now also firmly caught in the immediate attention span of many payments providers, often piggybacking on previous experience obtained in the financial instruments domain. In this Journal I would like to draw your attention in particular to the Transition section, where a number of operational plays for payments services providers are showcased. At Capco we have benefited greatly from our infrastructure project work for respected providers to further our thinking about the future of the payments industry. I trust that you will find this edition of our award-winning publication inspiring for your own insight and decision making process when facing payments challenges. Enjoy!
Rob Heyvaert, Founder, Chairman and CEO, Capco
Making payments work
It is quite remarkable how payments, be they instruments or mechanisms, are generally viewed, even by financial institutions, as nothing more than a means to an end. A service that needs to exist in order to make sure individuals and institutions are able to pay for the goods and services they purchase from one another. It is, if you will, the not so exciting part of the business. The bit that happens when all the other exciting events have taken place. Consequently, payments service providers, be they banks or other third-party institutions, have all to lose and not much to gain. If the transactions are settled smoothly then the mechanisms have done their job, but if there is any delay or problems, they have failed in this simple task. But, how simple is this task? Based on the articles in this issue, it is no mean task to make sure individuals and organizations who know little about each other can make payments to one another. The articles in this issue also highlight the fact that it takes more than smart technologies and goodwill from financial institutions to make a new payment mechanism work. We will find out why individual and institutional behaviors can make or break new payments instruments or mechanisms. And, more interestingly, why despite the fact that payments are of tremendous importance to financial institutions, in terms of costs and revenues, they remain so far below the radar of top management and strategists. However, before we discuss the contents of the three sections that make up this issue of the Journal, we would like to express our tremendous excitement, pride, and honor at the fact that the Nobel Laureate View for this issue has been kindly provided by Prof. John Nash Jr. of Princeton University. In section one, we focus on the topic of emerging payments models. Here we look at the new payment instruments and mechanisms that have been introduced and discuss why some have been more successful than others. The authors describe how new payment methods, such as debit, credit and, smart cards, are aiming to help individuals improve the way they manage their finances and make the whole process of payments more efficient. They also explain how card payments technologies are not only providing us with guidance on how to make our current payment methods more efficient, but how they can be essential when creating a whole new banking system for a war-torn country, such as Iraq. Section two discusses how payment systems have evolved in recent years. The authors look at the challenges faced by the banking sector in improving the efficiency of payments processes and explain how they were able to overcome them. This section also covers how simulation tools can help test new payment mechanisms without bringing the whole payments system under the risk of failure. Section three looks at the future of payments, including new payments tools and methodologies. The articles examine the potential benefits of focusing on those areas in which banks have specific core competencies and how outsourcing to other banks or third-party providers can help in that effort. The long list of new regulations introduced has dramatically increased the cost of being in this space, and the papers in this section look at what options are open to banks facing them. We hope that we have achieved our objective of providing a thorough review of the major issues impacting the payments world. We also hope that you enjoy reading this issue and continue to support the journal by submitting your best ideas to us.
On behalf of the board of editors
The Nobel Laureate view
Nash’s theories and their impact on economics and finance A discussion with Prof. John Nash, Jr. Senior Research Mathematician, Department of Mathematics, Princeton University, and Joint-Winner of The Bank of Sweden Prize in Economic Sciences in Memory of Alfred Nobel 1994
We are very honored that John Nash, Jr., the man credited
Q: Is that a quotation from your own speech at the awards
with one of the most profound contributions to the world of
ceremony?
economics, has kindly taken the time to help us get a better understanding of his theories and how they apply to the world
Prof. Nash: No, there is a misconception here! The fact is that,
of economics. Below, you will find the questions posed and the
quite exceptionally, I did not give a lecture in Stockholm on
answers kindly provided by Prof. Nash.
the occasion of my being awarded the prize (with Harsanyi and Selten). Instead, Prof. Harold Kuhn, of Princeton University, gave a lecture about the work I had done which was being
Understanding of the theories
recognized and about connections and applications, etc. This
Q: One of the most interesting facts about economic theo-
organizational format was arranged because I had not been,
ries that I have come across is the number of people, includ-
for many years prior to 1994, actually studying and working in
ing myself, who feel that they have a good understanding of
economics and game theory. And this circumstance was
a given theory when in reality they do not. Do you believe
related to my history of having been under ‘mental illness’ for
that most people who refer to your theories fully under-
many years, although at that time, in 1994, I could have, well
stand them?
enough, given a lecture myself, if that had been requested. So
Prof. Nash: No, indeed, the majority — even, say, among
on game theory or on economics. In years subsequent to 1994
while in Sweden I gave a lecture at Uppsala but this was not economists — would have only an incomplete or ‘casual’ famil-
and 1995 I have been, again, actively studying in these areas,
iarity with my contributions.
with new research projects.
Now I just happen to think of a very good reference source;
Q: In order to help our readers get a better understanding of
this is the book ‘Essays on Game Theory’ published by Edward
Prof. Nash’s contribution, we will use three quotes from the
Elgar (a British publisher specializing in economics and relat-
panel chaired by Prof. Kuhn in honor of Prof. Nash’s Nobel
ed fields). Elgar contacted me about publishing a book like
Prize in Sweden.
this soon after the news of my Nobel Laureate recognition. This book has everything that was published during the time
‘Nash proved by page 6 of his thesis that every n-person finite
of my studies in the 50’s in a major publication.
non-cooperative game has at least one (Nash) equilibrium
Q: Of course, many of our readers either do not have the
which is such that no player can improve his payoff by changing his mixed strategy unilaterally.’ Kuhn (1994)1
point. This is a profile of mixed strategies, one for each player, time to read that book or will have a tough time understanding its complexities. In order to help our readers get a better understanding of your theories can you kindly describe your
‘[The Nash] equilibrium is without doubt the single game
biggest contribution in a sentence or two?
theoretic solution concept that is most frequently applied in economics. Economic applications include oligopoly, entry
8 - The
Prof. Nash: It is not, really, a meaningful issue of whatever
and exit, market equilibrium, search, location, bargaining,
might seem to be my ‘biggest contribution’, when learning
product quality, auctions, insurance, principal-agent [prob-
and culture are viewed from some time much further into the
lems], higher education, discrimination, public goods, what
future. Right now, I can only go with conventional opinion and
have you. On the political front, applications include voting,
suggest simply a quotation of the citation at the time of the
arms control and inspection, as well as most international
Nobel ceremonies on 10 December 1994.
political models (deterrence, etc.). Biological applications all
Journal of financial transformation
1 Kuhn, H. W., 1994, “Nobel Seminar in honor of the work of John Nash in Game Theory,” December 8
deal with forms of strategic equilibrium; they suggest an
Von Neumann and Morgenstern’s book contains an excellent
interpretation of equilibrium quite different from the usual
mathematical analysis of one class of non-cooperative games,
overt rationalism. We cannot even begin to survey all of this
viz. of two-person zero-sum games and of the minimax solu-
literature here.’ Aumann (1987)2
tion for such games. It contains also an excellent mathematical discussion of one cooperative solution concept, that of stable sets, for many specific games. Yet, it so happens that the concept of two-person zero-sum games has very few real-
‘In the short period of 1950-53, John Nash published four brilliant papers3, in which he made at least three fundamentally
life applications outside of the military field. The concept of
important contributions to game theory:
stable sets has even fewer empirical applications.
1. He introduced the distinction between cooperative and
‘Had these two distinguished authors had Nash’s notions of
non-cooperative games. The former are games in which
cooperative and non-cooperative games available to them,
the players can make enforceable agreements and can
then presumably they would have asked the question of how
also make irrevocable threats to other players. That is to
to act rationally in a two-person nonzero-sum game or in a
say, they can fully commit themselves to specific strate-
more-than-two-person game if this is played as a non-cooper-
gies. In contrast, in non-cooperative games, such self-com-
ative game, permitting no enforceable agreements and no
mitment is not possible. (Actually, Nash also assumed that
irrevocable threats. Perhaps they would have asked also
in a non-cooperative game, the players will be unable to
whether one could not find for cooperative games a more
communicate with each other. Yet, in my own view, this
convincing solution concept than stable sets are. For instance,
would be a needlessly restrictive assumption. For if the
whether one could not find a solution concept yielding sharp-
players cannot enter into enforceable agreements, then
er-predictions about the players’ actual payoffs than the concept of stable sets does.’ Harsanyi (1994)4
their ability to communicate will be of no real help toward a cooperative outcome.) 2. As a natural solution concept for non-cooperative games,
Impact on financial markets
he introduced the concept of equilibrium points, now usu-
Economists view the markets, in specific stock markets, to be
ally described as Nash equilibria. He also established their
efficient, one of the main components of which being perfect
existence in all finite games. (Note that Nash equilibria
competition and perfect dissemination of information, such
seem to be the only solution concept applying both to
that no one can use their knowledge to gain superior returns.
games in normal form and in extensive form.) 3. As a solution concept for two-person cooperative games,
Q: You suggest that transactions fail not necessarily because
he proposed the Nash bargaining solution, first for games
participants behave irrationally, but because they do not
with fixed threats, and later also for games with variable
know each other’s perspectives of the transaction (i.e. what
threats. He also showed that, in the latter case, the two
is the highest price they are willing to buy at or the lowest
players’ optimal strategies will have maximin and minimax
price they will sell at). Does this mean that better informa-
properties.
tion can help increase the likelihood of transactions?
‘The best way to understand the importance of Nash’s contri-
Prof. Nash: I, personally, did not specifically suggest this. But
butions is by comparing the state of game theory just after
this phenomenon is a very real and recognized one. Recently
publication of Von Neumann and Morgenstern’s book in 1944
the theme of an (economics) Nobel Prize award (shared by
with its state after publication of Nash’s four papers in 1953.
three Americans) was ‘asymmetric information’ representing
2 Aumann, R. J. 1987, “Game theory, In The new Palgrave Dictionary of Economics,” edited by Milgate, M. and P. Newman, 460-482. This quote was extracted from the introductory speech given by Prof. Kuhn at the 1994 Nobel Prize seminar. 3 Nash, Jr., John F., 1950a, “Equilibrium points in n-person games,” Proceedings National Academy of Sciences, 36, 48-49 Nash, Jr., John F., 1950b, “The bargaining problem,” Econometrica, 18, 155-162
Nash, Jr., John F., 1951, “Non-cooperative games,” Annals of Mathematics, 54, 286-295 Nash, Jr., John F., 1953, “Two-person cooperative games,” Econometrica, 21, 128-140 4 Harsanyi, J.C., 1994, Nobel Seminar in honor of the work of John Nash in Game Theory, December 8
9
decisions of an individual negotiant. Learning is good in general, the interaction of speculators and investors is (of course)
Prof. Nash: I guess it must have been a long time ago that the
game-like, but beyond that truism the level of practical utility
stockbroker was invented. And through them occur purchase
may not be easily reached.
and sale transactions between parties that would be quite naturally very untrusting of each other. I would not want to
Q: Can your theories be used to help in determining man-
claim any magical potency of game theory here, but of course
agement compensation in the U.S., which seems to be going
it is naturally appropriate for the consideration of situations
through the roof?
where there are parties that have interests that are not simply parallel.
Prof. Nash: It is, indeed, a phenomenon. But, game theoretically, there is hardly any reasonable limit on the appropriate
Q: Would it help reduce the costs involved in international
compensation for a valuable executive. If Henry Ford, the
transactions?
founder, deserved to own the company, then at a later time, Bill Ford, a descendant and now chief executive (if any good
Prof. Nash: I feel that the costs of international transactions
at that function!), might merit a comparable level of compen-
are typically ‘what the market will bear’. Not long ago it was
sation. And there is the general phenomenon of the appar-
costly to change Italian lira money into French francs and vice
ently increasing disparity of incomes and prosperity. One
versa and a tourist could pay out a lot that way on a short
thing to think about is the increasing factor of multiplication
vacation. Now, with the euro that particular cost is entirely
applicable to the labor of a few persons. An actor on a stage
gone! Telephone charges have also typically been ‘what the
can be seen only by those in an audience within hearing dis-
market will bear’ rather than amounts actually directly relat-
tance. An actor in a TV series can be seen and heard by mil-
ed to costs calculable in terms of the technology.
lions at each performance.
World trade Payments
Founding fathers of economics have suggested that in certain
The theme of this issue is payments. One of the biggest prob-
circumstances even if the cost of production of one product
lems that exist in the world today is the lack of trust that
is higher in a given market it might still make more sense for
exists between market participants. Hence the need for the
that product to be made in that location if the cost of
establishment of ever more secure transactions.
exchange into other products is lower than in other markets.
Q: On payments, the inability to know about the motivations
Q: Can you explain why it is that most countries are still
and situations of many individual participants led to the
trying so hard to produce everything?
creation of exchanges and other mechanisms for central assumption of accountability for completion of transac-
Prof. Nash: There is often simply a national defense concept.
tions. The cost of learning the situation of participants is
The Japanese perhaps would feel themselves to be in an
often so high that other structures are created to reduce
unfortunate situation if they came to actually grow no rice at
the need to know. You could look at DVP as being refusal to
all with all rice consumed being imported.
learn anything: just pay me when I pay you and life is simple. Can non-cooperative game theory suggest a way to
Q: Do you believe that your theories can help improve the
improve the way organizations make payments to one
productivity of the world we live in?
another?
10 - The
Journal of financial transformation
the theme of how one party might be better (or differently)
one of the other, they are both contending with the auction-
informed than another. And this, of course, in itself implies the
eer.
importance of imperfections of information. The issues of
Q: Consequently, are you suggesting that investors/traders
completeness or incompleteness of information are old topics
should spend more time determining what the most efficient
in game theory.
level of negotiation would be?
Q: Can your theory also work when you are dealing with
Prof. Nash: Well, first of all, the concept of a ‘minmax’ or
thousands of participants? Can we really have behavioral
‘minimax’ is a concept relating to zero-sum games and thus
information about so many participants, or do we, can we,
properly relating to games of two players with directly oppos-
segregate them into groups to make this more viable?
ing interests. This is not really the relevant context for investors, because all investors can gain, in principle, with good
Prof. Nash: Most definitely yes! That is, the ‘non-cooperative
investments, or all could lose, in a great bubble of tulips or of
game’ concept was always, as I could see immediately, well
dotcoms. When the games of speculators can properly be
suited for application to the multitude of speculators and
seen as essentially of zero-sum character, with wealth merely
investors that might be trading in the shares of MSFT on the
being transferred but never created, then still the minmax
stock exchanges. You see, these individual traders are all act-
concept is not of use because these speculative games are
ing ‘incoherently’.
many-person in character (or even games of multitudes)
Q: Can the determination of other participants’ values of a
issues but relating to game theory in general I can make the
rather than games of two parties. And now aside from minmax transaction be helpful in increasing the returns we generate
general remark that a theory can be quite valuable even if it
from our transactions (i.e. are the people who are able to
does not give explicitly useful information to the persons of
master this able to consistently achieve better rates)?
the classes to which it would seem to directly apply. Thus game theory is not very helpful to chess players. And the
Prof. Nash: From the competitive point of view it is generally
medical and physiological theory of muscle functions and
desirable to have more relevant information. From the coop-
health would not directly help an athlete to prepare for the
erative point of view it can also improve mutual benefits for
Olympics. A player in games or a businessman or financier
the parties to have more information about themselves. There
may, realistically, need to use the relevant school of common
can be paradoxes, however, which can be revealed through
sense to play well, while game theorists may be able to effec-
game theory. I was surprised recently when I was at a game
tively analyze the behavior of classes of such players.
theory meeting and I heard a talk in which an example was presented of an auction situation, viewed through game theo-
J. P. Morgan said it all, in a sense: ‘Buy low, sell high.’ And this
ry, and it turned out that if two bidders in the auction had
is perfectly valid, although somewhat tautological.
different amounts of information about the quality of the product being sold at auction that it could happen that the
The ‘negotiant’ cannot usually rationally assume that other
bidder with less information would realize the higher payoff
negotiants are playing according to some sort of a game-the-
result! I was quite surprised when I heard this and I thought
oretically based strategy. If there is a ‘market’ then the advice
about it in relation to non-cooperative game theory. And then
of Morgan is relevant as a guide. So a game-theoretic analysis
I could see that it was consistent. Then a friend confirmed that
based on what others would be doing if they were operating
it was a familiar occurrence in examples. The thing is that the
according to a game equilibrium (or some other game con-
two bidders, in the auction game, are not simply opponents,
cept) might not be well justified, in an individual case for the
11
Prof. Nash: It would not be a direct effect if there would be a benefit. What I can see is that global cooperation in arts and sciences will naturally increase the economically measured ‘productivity’ concepts. Q: Many suggest that your theories could be used to help protect the environment in a way that benefits all. Why do you feel it has not been fully utilized? Prof. Nash: At the meeting in Marseilles devoted to game theory which I attended in early July 2004 I saw a poster by a Japanese theorist relating to the Kyoto protocols and the theme of global warming. The theme was developed in terms of cooperative theory but as I thought about it I felt that there would be a major issue deriving from the possibility of noncooperative behavior by nations with different natural interests. Perhaps in general, the environment, as an interest of all humans, can be most favorably or most easily protected by concerted human actions and planning if there is an appropriate and good level of cooperation of the world directed to that objective. Global cooperation, itself, is a natural topic for game theoretic studies. Q: Prof. Nash, thank you very much for giving us this time. We hope that our readers have benefited from your insight.
12 - The
Journal of financial transformation
Emerging models
Bringing broadband banking to Iraq Unifying the European payments landscape Credit card security on the Net: Where is it today? Broadening the value proposition around payments outsourcing Bringing payments together Payment systems and regulation Why has stored-value not caught on? Online payments systems for e-commerce Six smart moves when playing the smart card game
Bringing broadband banking to Iraq Kathleen Tyson-Quah Chief Executive, Granularity Ltd
sales executive confirmed their successful use of R-BGAN in
tions, however, I returned to Google. Applying the same crite-
three banking installations in Africa and the Middle East. In
ria of proven bank deployment and challenging project geo
response to my inquiry she let me know that NSSL was then
graphy, I was led to Pointsec Mobile Technology. Their solu-
supporting solutions to several clients operating in Iraq. At
tions, based on disk level encryption, had been deployed for
this point I asked whether NSSL would like to build a banking
thousands of bank laptops in the U.K. and U.S. and proven
network at 80 locations in Iraq. After a long pause, she sug-
secure worldwide.
gested that she would arrange a conference call with colleagues. NSSL has since confirmed that an internal debate on
I picked up the phone and dialed the U.K. representative.
whether I was too crazy to be taken seriously occurred prior
Again there was a pause and a sucking of breath when I sug-
to the conference call.
gested a rapid deployment to a network of laptops operating in Iraq. Then he said, ‘Sure. How soon do you want delivery?’
I must have acquitted myself rationally on the call, because
Another piece of the puzzle fell into place.
from that date NSSL was on my team. They justified my confidence in their project management skills by coming back
Now I had the network solution, I still needed the functionality
with a project specification and initial quotation for an
which would make it worthwhile. This time I did not need
80-node network within 48 hours.
Google. I knew who to call, Visa International. The choice was based on the belief that Visa could pop Iraqi Dinar on its
Because NSSL was already operating within Iraq, they were
National Net Settlement System as an additional currency and
able to simplify many of the logistics challenges. They agreed
process nationwide payments without major incremental
to take responsibility for procurement, configuration, and
investment of time and capital in project development. Using
deployment of the laptop PCs as well as on-site installation of
an existing, proven, high capacity system was essential to
the satellite modems. Their existing helpdesk operations and
meeting the tight timeframe for enabling the currency
Baghdad-based engineers would handle all downstream sup-
exchange.
port. NSSL would later assume responsibility for in-country training of hundreds of bank staff who had never before seen
Visa International’s National Net Settlement System operates
a computer mouse or used Microsoft applications, and also for
domestic settlements of card and electronic payment transac-
hosting the Central Bank of Iraq website.
tions for many countries, including countries in the Middle East and transformation economies in Eastern Europe. The
Mammoth mainframes spring to mind when payments are
system provides secure payment processes, backed by the
discussed among central bankers — not laptop PCs. Massive
VisaNet security safeguards essential to authorizing transac-
security and resiliency to safeguard the integrity of the pay-
tions and containing the risk of fraud.
ments process and payments data are essential. Laptops are viewed as inherently risky and insecure. I have read every
14 - The
Connecting to Visa would also have real advantages for inte-
Basle Committee report on payment systems and knew all
grating Iraq with the global economy, allowing Iraqis abroad to
this, but in Iraq it seemed to me that laptops have an edge
send money to Iraq for investment and enabling Iraqi busi-
from a resiliency perspective. If a laptop gets lost, stolen, or
nesses to make efficient payments to foreign suppliers.
blown up, we can put a fully configured replacement out with
Importantly, it would embed a global data standard, the
the next convoy.
16-digit card number, as part of payment operations in Iraq.
For state-of-the-art laptop security proven in banking applica-
I called Stuart Brocklehurst, Senior VP-Payments Development
Journal of financial transformation
One of the benefits of staying current with emerging tech-
I bounced the problem off a friend specializing in wireless
nologies is that it engenders a belief in the possible. Last year
telecoms and he immediately e-mailed back suggesting that I
I was challenged with the seemingly impossible, to design a
evaluate data services provided via satellite by Inmarsat.
banking network and payment system for post-war Iraq which
Some intense Googling led me to the Regional Broadband
could be fully operational in 10 weeks, and found that satellite
Global Area Network (R-BGAN). Portable R-BGAN modems,
data networks and web-based processes made it practical,
which are about the same size and shape as laptop PCs, pro-
quick, secure, and more functional than anyone expected.
vide broadband connectivity to the worldwide web via downlink to an earth station at Fucino, Italy. R-BGAN could be used
In July 2003 the telecoms were still knocked out in most of
to provide real-time access to any web-enabled application
central Iraq, the electricity nationwide was patchy, with fre-
operating remotely on servers outside Iraq. Crucially, Iraq is
quent blackouts, and only a handful of bank branches had
centrally located within the footprint of the Inmarsat satellite
reopened following widespread post-war looting of banks.
supporting R-BGAN. An added advantage of this web connec-
Meanwhile the security situation was deteriorating as insur-
tivity would be the ease of transferring services to a secure,
gents organized and gained access to weapons and explo-
domestic Internet Protocol network when it becomes available
sives.
in Iraq.
Despite these challenges, the Coalition Provisional Authority
Once I had broadband, I knew I could find a solution! I began
was determined to replace the old currency featuring Saddam
to sketch it out as laptop PCs connected by the R-BGAN
with a new currency. The changeover was scheduled to begin
modems to a back-end server providing the communications
on 15 October 2003, and last for three months. The banks
network, e-mail, and payments applications. Both the laptops
nationwide would have to reopen to operate the currency
and modems could operate off battery power during the fre-
exchange. In a country where many people keep boxes, even
quent blackouts.
rooms, full of cash in their homes in preference to the dubious security of banks, the exchange was going to be a major logis-
At this point I reined in my enthusiasm and got practical. I set
tical challenge for everyone.
myself some guidelines on technology and suppliers which accorded with the scale of the challenges in post-war Iraq. I
At an informal summer barbeque, Andrew Bailey (Executive
decided I would only consider suppliers offering technologies
Director — Banking and Chief Cashier) of the Bank of England,
proven in bank deployments in challenging geographies. Iraq
took me to the vegetable patch at the back of his garden and
was no place to test new technologies or unseasoned project
asked me to think about the problem. To say I was ‘led up the
managers. I also determined to keep as much of the back-end
garden path’ would be unfair, as he was frank in describing the
servers and operational processes as possible outside Iraq,
chaos. The urgent need was a network which would enable
although all processes should support the Arabic language
communications among 80 branches of the Central Bank of
and be transferable to a more settled domestic banking sys-
Iraq, Rafidain Bank, and Rasheed Bank so that they could
tem.
reopen throughout Iraq, and a system to process electronic payments among them.
Going back to Google, I kept searching on R-BGAN until I found a case study on deployment to major global banks at branches
I could have laughed away the request, but I felt I had to make
in the less developed regions of Africa. The deployment had
an effort to aid a country and a people attacked and invaded
been managed by NSSL, a satellite solutions company based
by my own.
conveniently in Redhill, England. When I first called NSSL, the
15
for Central Europe, the Middle East, and Africa. I explained the
I had invested a lot of myself in the solution, but the rest of me
challenge, the timescale, and the solution architecture I had
was happy to be sunning poolside at a villa in Greece, con-
developed so far, and asked if Visa would provide the pay-
nected to the team by an R-BGAN modem loaned to me for the
ments functionality. Stuart asked a few questions and prom-
purpose. They returned to London eight long, hot days later,
ised to get back to me within the hour.
the day before the bomb at U.N. headquarters in Baghdad which marked the most significant escalation in insurgent
A half-hour later he called to confirm the president’s view,
violence.
subject to approvals, that Visa could deliver. Over the following weekend all board directors approved Visa International’s
The solution went live on time and to budget on 15 October,
commitment to the Iraqi banking network and payment sys-
2003, pretty much as designed in my initial proposal. It was
tem. Given that we were in high holiday season, and many
enhanced by eight levels of security: physical, logical, network,
directors were tracked down to beach cottages and boats,
data, access, and more. Since then it has been used to coordi-
their unanimous support demonstrated a marvelous enthusi-
nate the banking operations of the central bank, Rafidain, and
asm. Visa even offered to cover the substantial set up costs in
Rasheed, bringing the banking system back on-line and mov-
Iraq.
ing it toward normal operations.
Connecting Visa to the Iraqi banking network was not the
This summer, 2004, the system embraced the 26 private
technology ‘no-brainer’ I had hoped. Visa uses a proprietary
banks operating in Iraq, enabling them to offer Visa transac-
network and data standard. NSSL quickly went to work with
tions and make settlements through the central bank and
Visa’s network supplier to overcome this, enabling the trans-
globally.
mission of transaction data from Iraq over Internet protocol using a web-based connection and data transformation for
There is still a long, long way to go before Iraq becomes a
interoperability with the Visa network. It was an impressive
modern economy with a modern banking system. The solution
cooperation, displaying a ‘Blitz spirit’ toward engineering a
we designed and deployed was aimed at meeting an urgent
rapid, workable solution.
need in difficult times. It provides a starting point on a long path of modernizing the Iraqi banking system.
My initial collaborative proposal was sent to the CPA in Baghdad the following week. The CPA had not really expected
All solutions to complex problems look simple in retrospect.
anyone to solve the problem, so was a bit surprised to receive
The solution deployed in Iraq may appear obvious revealed as
a fully-scoped solution. It confronted them with another prob-
a completed whole, but there was twenty years of applied
lem. They could not award a contract without a tender pro-
expertise on what a payment system is, how it needs to oper-
cess, so the Iraqi Banking Network and Payment System had
ate, how banks must interact, and how the processes must be
to be put out for bid. Fourteen companies responded, but no
protected which went into specifying the solution last sum-
other proposal provided high security and resiliency with
mer. Combining payments expertise with workable technology
equivalent confidence in delivery on the timescale required.
and suppliers who could deliver in Iraq gave our solution cred-
Ours was also the lowest cost.
ibility with the bankers responsible for using it in daily operations.
In the first week of August 2003, Stuart and two executives
16 - The
from NSSL flew into Baghdad in an RAF Hercules transport to
The Iraqi banking network and payment system can be count-
negotiate the contracts. Part of me wanted to be with them as
ed among the few successes of the past year in Iraq. Even so,
Journal of financial transformation
I learned some valuable lessons that have nothing to do with technology or banking. First lesson: no contract, no client. In my zeal to get the job done, self-interest took a back seat. Next time I would sort some agreements to ensure I get paid. Second, corruption can frustrate the best technology, and the Iraqi banking sector has been a magnet for returning exile carpetbaggers. Even with the overwhelming support of highly professional and enthusiastic native Iraqis who are determined to rebuild their nation, our system was nearly derailed by the malign influence of a few interlopers. Finally, there is nothing more permanent than a temporary solution. Our network, originally to be replaced by a terrestrial system within 6 months, is still operating to connect the 81 branches installed on it, and NSSL and Visa are cooperating on a worldwide network leveraging the superior resiliency and security features developed for Iraq. It was a mad, exhilarating summer project. I have nothing but respect for the people in the CPA, Central Bank of Iraq, NSSL, and Visa International who made a reality of the ideas I spun from my desktop. A strong banking sector is essential to any economic normalization in Iraq, and we should all wish them success going forward.
17
Unifying the European payments landscape Philippe Menier Chief Operating Officer, Visa Europe Ltd example, Santander Central Hispano’s acquisition of Abbey
level. It is a bit like the famous old quote from St Augustine:
appears to be heading for a successful conclusion. And, with
‘Lord, help me to be good — but not just yet.’
every passing week there is more speculation about yet another mega-merger. Clearly, part of the business rationale
Now we seem, finally, to have reached the moment of truth and
is to achieve more scale and scope — optimizing the cost base,
the mindset of the industry does appear to be shifting. Even so,
and eliminating duplicate infrastructures.
we are still at the start of a very long, difficult process, and there are some very complex, emotive issues to resolve.
As a related point, there is the fate of the shared, domestic payment processing infrastructures. Everyone agrees that
The future of processing study
Europe cannot support a parallel infrastructure in every single
Against this background, Visa Europe commissioned Payment
country, so these incumbents are actively hunting for a new
Systems Europe to undertake a major research study among
remit. In the world of payment card processing, for example,
European banks and processors — and some readers may have
Interpay, Banksys, and SSB have pooled their scale by creating
participated in this. We asked for their opinions on today’s
the SiNSYS joint venture. Then of course there’s the SEPA
card processing market. We asked them how they thought it
vision itself — all those ring-fenced national silos to be super-
may evolve. The findings make for very interesting reading.
seded by one big domestic market. In terms of the big picture, there are probably not too many The banking industry may have been reluctant to change. But
surprises. Yes, everyone agrees that SEPA is a worthy aim.
the first round of regulation and the formation of the
And yes, everyone agrees that it makes sense to move
European Payments Council (EPC) proved that this really is
towards a common payments architecture. In other words,
not going to be a matter of choice. Coinciding with this we
there is complete consensus on what and why. Where it gets
have the emergence of a new legal framework for payments,
more tricky is when we discuss who, how, when, and where.
the Commission’s scheme to harmonize the legislative context
They say the devil is in the detail, and you do not have to go
for all payments, not just across the euro-zone, but all E.U.
very far before you meet him head on!
countries. Covering transparency and liability, security and fraud prevention, and revocability and refunds, it will create a
But the important thing to acknowledge is this: no one is
level playing field across the continent.
happy with the status quo. Not the banks, not the commercial processors, and not even the inter-bank processors. Even if
Looking more closely at our own sector, card payments, there
SEPA were not on the agenda, there would be real pressure to
is the European migration to EMV chip technology. This is
change, and to move toward more openness, transparency,
significant for many reasons but in this context it is the first
standardization, and competition. Indeed the study shows us
time that Europe is uniting around a single card standard. The
that it is not even SEPA that is the biggest force for change.
casual observer may think of the card market as highly stan-
Nor is it the Commission’s competition intervention. Instead it
dardized. But, scratch below the surface and you see a mass
is the needs and demands of the big pan-European banks
of national idiosyncrasies. EMV represents a big step towards
which is regarded as the primary change driver.
convergence. With regard to the banks, and looking at issuer processing in
18 - The
So, yes, many different examples of convergence. I suppose
particular, a real dilemma is revealed; namely, what, when, and
we have been using this word ‘convergence’ for years now —
how much to outsource. For many banks, it seems to be con-
toying with the possibilities, but mostly on a hypothetical
stantly on the agenda. For example, one of the 20 large banks
Journal of financial transformation
As many readers will be aware, one of the central objectives
Our perspective
of the European Union has been to create a single European
It should be noted from the onset that we are not a single
market where there are no barriers to trade or commerce
entity, rather the Visa organization, like our peers, is an inter-
across any of the member states. Although the European
national association of banks, uniting more than 5,000 banks
Union has made huge strides in this direction, several obsta-
in Europe, and 21,000 worldwide. In Europe, these member
cles continue to stand in the way of the vision, in particular,
banks have issued 230 million Visa cards which, during 2003,
cross-border retail payments have been seen as unduly slow
were used to conduct 824 billion worth of transactions, rep-
and expensive.
resenting close to ten percent of personal consumer expenditure across Europe. Member banks also sign up merchants to
A number of commentators have enjoyed pointing the finger
accept Visa, or other cards issued by our peers, in order to
of blame at Europe’s retail banks, but the fact of the matter is
benefit from fast, efficient, guaranteed cashless payments. We
that E.U. countries (just like countries everywhere else in the
act as the glue which binds everything together. For example,
world) have independently created a patchwork of parallel,
we manage the brand, provide the processing infrastructure,
incompatible payment networks. Given that the vast majority
and work with banks and regulators to agree and enforce the
of payments has always been, and continues to be, domestic,
operating rules and regulations, defining the precise way that
there has never been a compelling business case for bringing
payments must work, and the responsibilities of each partici-
more unity to the national infrastructures. The net result is
pant in every transaction.
that, yes, cross-border payments have been slow and expensive. And this is generally reflected in the price and the qual-
European banks were the first non-American participants in
ity of service experienced by businesses and consumers alike.
Visa. They were also the first to regard it, not as a credit card
The one notable exception is the card payment sector. Since
system, but a generic payment mechanism. Accordingly, they
its inception, the card systems have effectively ‘overlaid’ a
were the first to issue Visa debit cards, and the vast majority
series of domestic payment schemes with international func-
of our European cards continue to be debit. Since these debit
tionality and interoperability. In doing so, it has enabled
cards are the dominant payment means and the pivotal point
highly convenient, highly secure, generally instantaneous
of the banking relationship, they tend to have real strategic
cross-border and multi-currency payments. And it has done
significance for any retail bank. Accordingly, effective gover-
so at minimal cost.
nance and control of the payment system is a top priority for European banks, and a highly pertinent issue for European
Frustrated at the lack of progress across other types of retail
regulators. It is, therefore, worth admitting to a vested inter-
payments, the European Commission is taking a tougher
est — by acknowledging that Visa has a very big stake in the
stance. It has already introduced payment-related legislation
way the European payments industry evolves. And, on behalf
and has insisted that Europe’s banking community must now
of our member banks, we are working hard to influence that
progress towards a Single European Payments Area (SEPA),
evolution.
where any regional payments can be made as easily and cheaply as in the local market. Given the success of the pay-
Convergence across banking and payments
ments card industry in mitigating, if not eliminating, a num-
Before looking at the specifics of the European retail pay-
ber of the aforementioned issues, it might be worth assessing
ments sector, it is perhaps worth considering the broader
whether the card payment sector, either directly or by exam-
context, by which I mean the multiple convergence trends
ple, can play a role in helping the industry achieve its objec-
across the banking and payments sectors. Firstly, the mergers
tives.
and acquisitions continue apace. At the time of writing, for
19
cult situation. They say that the banks have a new-found pre-
was recognized that the scale and the functionality of VisaNet
occupation with costs and efficiencies, often at the expense of
could add value to banks and the processing infrastructures
revenue generation and profitability. Consequently, all they
they are using today. Consequently most banks had accepted
want from the processors is vanilla-style commodity services.
that, yes, they would have to invest — at some stage. But with a lack of clarity regarding the future landscape, it was gener-
Clearly, it is not at all easy for any processor to excel in these
ally thought that now was not the time to make such an
circumstances, particularly when you have to upgrade your
investment. It was also thought that, yes, in an ideal world it
systems, the market is highly oversupplied, and there are bar-
would be best to have a choice of competitive PEACHes. But,
riers to international growth.
of course, this is not an ideal world, and there was no appetite to co-invest in an alternative PEACH infrastructure.
Ultimately, one is drawn to the conclusion that, with 80-plus commercial and inter-bank processors something has to give.
Finally, we asked these large banks about the environment ten
The banks said, yes, the market has to shrink substantially,
years from now. They all agreed on two points. Firstly, they all
and the consensus was that there may be room for 20 to 30
agreed that, ten years from now, there would not be more
players. The processors agreed, they too said there was room
than 4 to 5 wholesale payment clearers. And, secondly, all 12
for 20 to 30 players and suggested that would be the size of
banks we spoke to agreed that they would be one of the four
the market just seven years from now.
or five survivors!
Other cross-border retail payments
Where to from here?
It should be stressed that what we have been discussing so far
Even if we are to confine our considerations to the payment
is the payment card market and payment card processing.
card sector, it is clear that Europe’s banks have some difficult
This is a payments market which is commonly perceived as
choices to make. Let us look at some of the criteria that I
the most uniform, the one that is closest to the SEPA ideals,
believe will be important in the decision-making process.
and has the fewest adjustments to make. So what about the prospects for other types of payment?
Grasping the EMV opportunity
Visa has also spent some time analyzing the ACH market.
that the European credit card market functions, and this also
Specifically, we conducted some very detailed, structured
extends to charge cards and some types of debit card. This,
discussions with our members, and also with the national
however, is not the case for a large chunk of the debit market,
As I have discussed, there is definite consistency in the way
ACHs. Incidentally, our aim was not to define how Visa could
particularly in central Europe. Indeed, there are 15 different
replace existing infrastructures and market players, rather, we
domestic card implementations, each of which has its own
wanted to examine whether our processing capabilities could
particular rules, regulations, and processing infrastructures.
enhance the capabilities of incumbent players — perhaps by
The casual observer could suggest that the credit card model
providing complementary, value-added processing services,
should simply be exported across Europe. Given the history,
which could result in reduced costs and increased revenues
and the fact that the debit card is so very pivotal in the bank-
for banks.
ing relationship, the industry insider would label such a suggestion as naïve.
Again, the findings are extremely interesting. Firstly, it was
20 - The
thought inevitable that pan-European clearing infrastructures
However, with the implementation of EMV chip technology
will emerge to replace the existing, local ACHs. Moreover, it
right across the region, there is an historic opportunity to
Journal of financial transformation
we spoke to said they would outsource all of their issuer pro-
big pan-European banks really need is multi-country solutions
cessing immediately, if only a suitable processor were avail-
for issuing, acquiring, and ATMs. Today, few if any processors
able. Another one had just changed its processor. Another two
could offer a credible service.
were in the process of changing, and yet another had just moved its processing back in-house.
So, how about the processors themselves? These organizations fall into two broad camps, the commercial processors, or
So what is going on here? These banks say they are frustrated
TPPs, many of which are owned and/or operated by large
by the lack of choice and flexibility, they do not like to be
American multinationals, and the mutually owned inter-bank
dominated by inter-bank processors, and they are not
processors. It is interesting to note that, although the U.S.-
impressed with what is on offer from the third-party proces-
style third-party processors are admired for their commercial
sors either. Yes, they have got an element of choice when it
focus they do not seem to have adjusted to the realities of
comes to their credit card processing, but, for banks in main-
Europe. Banks say they are unhappy with their style and com-
land Europe, the priority is invariably debit.
plain that the solutions they are offered tend to be based on the U.S. or Anglo-Saxon model. They are frustrated by a lack
But how do they break the deadlock? Again, it is agreed that
of understanding of the specific requirements of individual
the larger banks will be the ones to decide. On balance these
banks in different European countries. It is even suggested
big banks say that, yes, they would like to see closed domestic
that by simply getting out on the road more often senior
markets opened up to competition. They are in a situation
executives could do much to address the skepticism. Also,
where they are supporting a different inter-bank scheme in
thanks to their file conversion expertise, they have a particu-
every country where they operate, and each scheme is pre-
lar opportunity in cross-border processing. And, as I said
senting them with a big bill to replace an aging delivery plat-
before, the bigger banks are actively looking for such solu-
form.
tions.
But there are not any easy exit strategies. One option could be
I should stress that this is not an exercise in anti-Americanism.
to renew their support for these incumbents, helping them to
Many of our home grown processors were severely chastised.
pursue pan-European strategies, to win cross-border volumes
Banks tend to believe they get higher quality services from
and thereby drive down cost. Alternatively, they could orches-
their in-house operations, and this reflects badly on the entire
trate the sale of these inter-bank processors, dismantling local
sector.
defensive barriers, and effectively opening the door to the U.S.-style third-party processors. Both strategies have definite
Moving to the inter-bank processors, the survey suggests that
drawbacks, so the more likely scenario is gradual, market-by-
they may be facing an uncertain future. There was some criti-
market liberalization, as we have begun to see in Scandinavia.
cism over the quality of service and the costs of redevelop-
Barriers could gradually be lifted, and step-by-step commer-
ment. However, no one should underestimate the level of
cial processors could be invited and encouraged to compete
support they retain. Several banks believe there is definitely a
with the inter-bank consortia. In this way, competition would
future for cooperative platform development and shared ser-
begin to change the market dynamics and improve the quality
vices. The conclusion is clear, so long as it adapts to the reali-
and range of services.
ties of a larger, more uniform European market, the inter-bank cooperation model is alive and well. It should also be noted
For many, this is a more palatable option. But, even so, it is
that many of the processors, irrespective of whether they are
slow, unpredictable, and still not a complete solution. What the
third parties or inter-bank consortia, face an impossibly diffi-
21
unite, and possibly even converge, around a common plat-
are also quite unique. Additionally, no other banks face the
form. What we are talking about here is the creation of a new
challenge of migrating from separate, multi-country infra-
Always Chip — Always PIN debit product for Europe. In a few
structures to a more unified regional infrastructure. The
years, Europe will be a mature chip market — indeed, more
recent constitutional changes at Visa Europe are a reflection
than 90% of terminals and cards will be EMV. In other words,
of this situation. Traditionally, we operated as a branch of
the environment will accommodate fully secure cards, debit
Visa International. But, as of July 2004, we created a sepa-
cards in particular — for which consumers’ security expecta-
rately incorporated company called Visa Europe. In effect,
tions are not fully met today, especially when they travel out-
this means that Europe’s banks are the joint-owners of a dis-
side of Europe. To the extent that there is a consensus, it
tinct European payments business, and to reflect their propri-
appears to be that Europe’s banks should aim for a fully
etary interest each one holds a nominal, non-transferable
secure debit solution, which at the point-of-sale will be chip
share in the company. Yet, to maintain our internationalism,
and PIN only.
Visa Europe exists as a group member — and co-owner — of Visa International.
With no exception items, no skimming of the magstripe, no fraud, and the opportunity to operate off-line in selected envi-
In business terms, incorporation means that we can accelerate
ronments, such a product would result in lower unit costs for
our European development work. It also means we can devel-
all participants, which would in turn lead to increased mer-
op more technology services and have the freedom and
chant and consumer confidence. Our own proposal is for a
opportunity to enter into commercial partnerships and joint
product which is fully owned and controlled by Europe’s
ventures. Incorporation also has profound symbolic signifi-
banks, and which operates under its own acceptance mark. It
cance. It demonstrates that Europe’s banks are the exclusive
would be able to coexist with domestic debit schemes, as long
owner-operators of their own regional payment system. With
as the banks who own these schemes find it appropriate. Of
an unambiguous European focus, and pan-European repre-
course, it would also be an ideal point-of-convergence for
sentation, we can express their views and represent their
these schemes. Although there are differing views about the
interests more forcefully with European associations, regula-
timing for such convergence, within and among countries, no
tory bodies, and forums. In particular, we want regulators,
one is arguing with the fact that it will and must happen at
industry bodies, and working groups to regard us as a wel-
some point. At Visa Europe, we are, therefore, eager to work
come and authoritative partner.
with each of our members to see how such a product could help them evolve towards a more secure, more convenient,
This issue of ownership and control also extends to the pro-
and more cost-efficient European debit scheme, topped up by
cessing infrastructure which ties together the various partici-
a European-only governance structure.
pants. Our various surveys suggest that Europe’s banks are unwilling to cede control to third-party technology companies.
Maintaining European ownership and control
Perhaps they have glanced across the Atlantic and seen the
What some observers fail to appreciate is that ownership and
weight of influence which such companies have on the terms
control of the card payment system is a highly emotive sub-
and the costs of the entire payment market.
ject for European banks. Thinking back to chip cards, for
22 - The
example, the EMV specification itself was hammered out in
In summary, I believe that Europe’s banks have a deep and
Europe, and Europe is the clear world leader in implementing
enduring appreciation of the strategic significance of elec-
it. The nature of the cards issued here in Europe is very dif-
tronic payments. I contend, therefore, that any companies or
ferent from elsewhere in the world, and their usage patterns
organizations which seek to represent or serve these banks
Journal of financial transformation
Credit card security on the Net: Where is it today? Sankarson Banerjee Senior Architect, Mphasis Internet commerce is here to stay. It has become a common
available to him or her at the point of purchase.
method of purchase for a wide range of products, and continues to grow at a good clip. The dominant payment vehicle on
Though the technology has been available for the better part
the Internet is, and has always been, the credit card. Indeed,
of a decade now, implementation even today is patchy. Credit
the two are inseparable. E-commerce does not do well where
card companies will not refuse transactions where the CVV
credit card usage is not prevalent, and the Internet has cer-
does not match; only a flag is sent back and the merchant is
tainly given a boost to card spending.
allowed to decide if CVV match should be enforced. More importantly, except in the case of American Express, collecting
When Internet commerce first came along, credit card issuers
and matching a CVV number does not provide a merchant with
were woefully under-equipped to handle the security issues
any additional protection against charge-backs. A purchase
that emerged from this new way of doing business. Credit
can be disputed and charged back to the merchant even if the
cards were meant to be physically presented at the point of
CVV number was correct. The rate of charge-back is lower than
the sale, so that the merchant could validate the transaction
the rates at which CVV matches fail, so even legitimate card-
and obtain a signature that represented a legally binding con-
holders are entering their CVV wrong some of the time.
tract to pay between cardholder and merchant, with the credit card intermediaries acting as facilitators. The Internet
CVV does not really prevent trusted party fraud. The waiter or
introduced a whole new paradigm, where sellers and buyers
grocery store clerk entrusted with swiping the card can copy
did not come face-to-face, could easily hide their identities,
the CVV just as easily as the card number. It eliminates the
and required no signature. This led to high rates of fraud that
situations where a discarded charge-slip or statement can be
often left merchants reeling and customers unhappy.
used to locate the credit card number, although getting credit card details from the former is already becoming more diffi-
The credit card industry has responded with many measures
cult as most now mask part or all of the number. It does,
aimed at lowering the rate of fraud to acceptable levels. Some
however, seem to lower the percentage of fraud; the effect is
measures have, over the years, become quite widespread. This
possibly that of a psychological deterrent.
article studies five schemes that were launched by the credit card industry in the last few years, and how well they have
3D Security
served their purpose.
3D, or Three Domain security, is an Authenticated Payment
Card Verification Value (CVV2)
authenticated, as the holder of the credit card, in addition to
Solution. This means that the person making the payment is One of the earliest measures dates back nearly a decade.
the payment itself being authorized. Visa’s solution is called
Also called the Card Validation Code (CVC2) or Card
‘Verified by Visa’. The way it works is that Visa asks the cus-
Identification Number (CID), it is a 3-digit number (4-digit for
tomer at a participating merchant’s site to confirm both the
Amex) printed on the credit card that is not in raised lettering
credit card number and enter a pre-decided PIN or password.
and not present in the magnetic strip. It does not show up on
If the PIN is not entered correctly, the transaction is refused.
credit card receipts either. These numbers are present in Visa,
It is very unlikely that a credit card thief will know the PIN, so
MasterCard, Discover, Amex, and Diners cards. MasterCard
the transaction should be safer. Further, the PIN cannot be
has had it on every card issued 1997. The way it works is
accessed by someone in temporary possession of the card.
simple; websites ask the user for the CVV2 along with the
MasterCard has a similar program, called SecureCode.
card number. It does not verify the identity of the cardholder; it just tries to ensure that the person has the card physically
The success of the system is dependent on two factors, that
23
the credit card customers will enroll themselves in the Verified
Address verification requires substantial investment by credit
for Visa program and create a PIN for themselves, and that
card issuers. Issuers are the only persons with access to the
the merchants will participate in the program. Customer
customer’s billing address and their ability to publish this
acceptance does not seem to have been poor; Visa estimates
information real-time to payment processors is critical to the
that nearly 40% of its customers in the U.S. are now enrolled.
AVS method. Countries outside the U.S. have been slow to
Merchants, however, have been far less enthusiastic. Barely a
adapt to AVS, so this method is usually not available for
handful of Internet sites support the mechanism, and most
International customers. Billing address even within the U.S. is
major retailers have not shown enthusiasm.
not guaranteed reliable. Some users do not update their addresses with the credit card company when they move, rely-
The question here is the kind of protection this program
ing instead on forwarding services of the post office. By and
affords, and who it protects. Assuming that the cardholder
large, however, billing addresses tend to be current and using
takes reasonable care with the PIN, this affords protection
them as verification can be quite effective.
against misuse on the sites that support it. However, two points must be made. Firstly, most consumers are already well
There are some problems, however. For example, credit card
protected from unauthorized charges on their credit cards by
thieves do not find it so difficult to locate the billing address
the card issuer. Except for the inconvenience of having to
of a person; these are routinely stored with the merchant
dispute a charge, the cardholder is usually not liable for a
along with the credit card number. It can also be obtained
fraudulent transaction. Secondly, unless the card covers a
from a person’s discarded statements. In this respect, it is
large majority of merchants everywhere on the Internet, there
much weaker than 3D Secure. Some of these problems can be
is very little to deter a card thief. The reason why merchants
addressed by insisting that goods (especially high value
have not shown enthusiasm is that they stand to gain little by
goods) be shipped only to the billing address, thus preventing
this program. Adopting the program does not get them lower
the thief from benefiting easily from the purchases. However,
interchange fees reflecting the lowered risk of doing business.
this creates more issues. It does not work for soft products,
It does not even protect them against the cost of bearing
such as software licenses or porn, and makes it more difficult
charge-backs. Customers can still dispute charges and, just as
for people to send gifts, which form a large percentage of the
in the case of CVV2, no extra consideration is given to mer-
commerce. Additionally, many people send their credit card
chants. Most merchants, therefore, are reluctant to make the
statements to their own or their accountant’s office. Because
change, which means implementation cost as well as custom-
of these issues, AVS is used more as a check than a mecha-
er issues (1-click ordering, for instance, will no longer work).
nism to guarantee a transaction.
Address Verification Systems (AVS)
Surrogate numbers
The theory behind address verification is that the billing
One clever way to avoid exposing credit card numbers to a
address of a credit card is more difficult for a thief to gather.
merchant is to use disposable credit card numbers generated
A waiter at a restaurant can easily note down the card number,
for a specific transaction and valid only for a fixed duration or
but has no way of obtaining the billing address. Furthermore,
with a particular merchant.
merchants can insist that high value products be shipped only
24 - The
to the billing address, making it less likely that someone other
Two companies lead in this technology, which has been
than the cardholder can benefit from the purchase. Address
adopted by some of the world’s largest card issuers. American
Verification is today available for all cardholders in the U.S.,
Express calls it Private Payments and MBNA calls it ShopSafe.
and in some other countries like the U.K.
The process is simple. When making a purchase, the customer
Journal of financial transformation
uses a small application to generate a one-use credit card
troublesome. While card reader hardware is often distributed
number with a fixed value and a short validity period (usually
free, few users go through the trouble of hooking them up and
two months). This card number is valid only for one transac-
installing them. The readers are inconvenient for laptop users,
tion, or for repeated transactions within the validity period.
who may not even have a slot to plug them into. Furthermore,
The credit card company maps these temporary numbers
most households in America have multiple computers, and
back to the cardholder’s original card.
usually only one reader is supplied for free with every card. Because of these and other issues, usage of the smart capa-
The technology is very effective when used against unknown
bilities of these credit cards is very low. Most people just use
merchants, especially on auction sites or for purchasing one-
them as regular credit cards. Merchants too do not invest in
off services from small sites. It is pretty much impossible for
fully smart-card enabled purchasing mechanisms.
the merchant to defraud on the transaction outside the scope defined by the customer.
Conclusion None of the new measures studied really provide the silver
Issues still remain. This approach negates the huge conve-
bullet, the one way to ensure a completely safe transaction on
nience of 1-click shopping, requiring the customer to fill the
the Internet for both merchants and consumers. Most offer
card details each time. Furthermore, for services which are
some level of increased protection, but do not really address
booked with a card but require the production of that card at
the issue of excessive charge-backs, an issue of primary con-
the time of taking delivery, such as with movie tickets or air
cern with merchants and issuers. This is directly reflected in
tickets, this approach causes significant headaches.
the reluctance by VISA and MasterCard to lower the interchange fees charged for Internet transactions even if one of
Another issue is that if this method is adopted for widespread
the specified methods is adopted.
use, there will soon be a shortage of numbers. Of the 16 digits on a credit card, only 10 are truly variable for a single issuer,
Credit card usage on the Internet is clearly on the way up.
leading to approximately 10 billion possible cards. That is not
Fraud rates, though higher than with other channels, have
a problem when physical cards are being issued, but it is cer-
held steady for some time. Much of this is the psychological
tainly not a large limit if each transaction uses a different
assurance that these measures, often well publicized, have
temporary card number.
contributed to better security. It increases consumer confidence in the system and promotes the perception that fraud
Chip cards and smart cards
is difficult.
Credit cards with embedded chips were supposed to be the ultimate security mechanism. Such cards are very common in
A more important reason is probably the increased ability of
Europe, where a large number of MasterCard branded credit
merchants to track and prosecute the big offenders, and their
cards have them. In the U.S. American Express is the only
increased ability to detect and predict fraudulent behavior so
major issuer to offer them, on one of their cards (The Blue
that they can carry out more detailed checks (possibly offline)
card).
on individual transactions that seem suspicious. That is prob-
If all the pieces of the credit card transaction were smart-card
detection of possible fraud, and confidence building via indus-
enabled, the transaction could indeed be made very secure.
try-wide measures.
ably the way it is going to continue — better policing, smarter
This, however, involves the use of a smart card reader and a computer with the appropriate software; both prove to be
25
Broadening the value proposition around payments outsourcing Thomas Halpin Vice President, JPMorgan Treasury Services advice-to-receive MT 210 matching tools that improve trea-
core capability and competitive advantage and consider the
sury and receivables management, as well as Web-based
outsourcing option from a broadened perspective.
tools that improve customer service and information reporting (i.e. faster inquiry response and the ability to proactively stem inquiries). Such capabilities help the buyers of an outsourcing solution to differentiate their payment services with their own customer base. Institutions considering migrating to an outsourcing solution should learn about the value-added capabilities that potential outsourcing providers include in their solutions. Such features help broaden an outsourcing solution’s value proposition to one that potentially improves client satisfaction and creates new revenue streams that support a model of business growth. At the same time, a buyer of payments outsourcing retains its direct clearing membership for continued visibility in the market.
Core competency or competitive advantage? Distinguishing between core competency and competitive advantage is central to the decision of whether to purchase an outsourcing solution. Potential buyers should ask, ‘How can we obtain competitive advantage in services that our clients expect us to deliver?’ Core competency is something an institution does well, but competitive advantage is what a firm must establish to differentiate itself in the market and create long-term sustainable wealth. The new reality for the financial services industry includes industry consolidation, new regulatory requirements, increasing business-resiliency measures, evolving client needs, and changing technology. Today financial institutions must honestly assess their business strategy to ensure they are directing resources in a manner optimal for generating sustainable profitability. When considering outsourcing payments, institutions should assess the impact of maintaining their current model and evaluate the future direction of the market and the changes required to stay competitive. They should distinguish between
26 - The
Journal of financial transformation
Historically, payments outsourcing has been viewed as a cost-
— in addition to hosting the technology platform.
savings measure. Today, financial institutions are searching for alternatives to maintaining their own infrastructures for
The key to a BSP approach is that the solution never limits the
delivering transaction services. Institutions are realizing that
client’s ability to control its payment flows. Rather, it gives the
a payments outsourcing solution has a broader value proposi-
client options and choices for how to manage their funds
tion than cost savings alone. The right solution can be a stra-
transfer activity. For example, the client can elect to handle
tegic decision that drives revenue growth, deepens client
the payment release, repair, and settlement, or choose to
relationships, and supports sustained profitability.
engage in balance monitoring to ensure that their customers have enough funds to authorize payment release. The buyer
Key to this broadened value proposition is finding the right
of the outsourcing service maintains control over the business
outsourcing provider, one with the scale to drive efficiency
processes and decides its level of involvement in the continu-
alongside robust product capabilities and best practices that
um of processes.
improve quality while providing superior controls. Further more, a provider should demonstrate a long-term commit-
Potential buyers of outsourcing services should consider
ment to maintaining a best-of-breed capability, which helps
whether a vendor offers both ASP and BSP models or pro-
the solution’s buyers gain a competitive edge in delivering
vides a technology platform only (ASP) and outsources the
payments services to their own customer base.
BSP component, as a vendor that acts as an application service provider, but also offers business process services, may
Tipping the scale towards profitable growth
be outsourcing that component to yet another provider.
Institutions without the scale of business to compete efficiently must find viable alternatives to their own payments
Maintaining excellence for the long-term
infrastructure. Many are looking for ways to leverage the
Buyers of payments outsourcing should be sure that potential
scale of other providers in order to stay in the payments
providers have a vested long-term interest in maintaining a
space. An outsourcing model is emerging as a strategic solu-
leading capability that incorporates best practice. This level of
tion that enables institutions to continue serving their mar-
commitment can be a differentiator when distinguishing
kets with payments services. And the right outsourcing pro-
among outsourcing providers. Demonstrated market leader-
vider can help its clients increase their existing capabilities so
ship often reflects a clear long-term commitment. When the
as to differentiate their services for competitive advantage.
underlying payments business is mission-critical to a provider’s overall franchise, it means that the provider has a vested
Choosing an outsourcing model
interest in maintaining a best-practice environment and stay-
A number of today’s outsourcing vendors, including banks,
ing at the forefront of the payment industry’s future direction.
serve as application service providers (ASP). A more select few also offer a business process services model (BSP). With
Consequently, best-in-class providers should enable clients to
the ASP model, the vendor hosts a funds transfer technology
benefit from straight-through-processing rates of close to 97
solution on behalf of the institutional client, while the client
percent on their pre-processing activities, which lowers the
continues to provide the resources for pre-processing and
number of process repairs. Of course, a lower error rate and
post-transaction service activities. With the BSP model, the
improved payment processing efficiency translates into more
vendor provides transaction processing and service activity
timely payments and improved liquidity management. Best-
— including investigations, repairs, and other day-to-day
in-class providers should be expected to have invested in
operations that support the client’s payment clearing activity
value-added features like e-mail notification of payments and
27
Bringing payments together Michel Akkermans Chairman and CEO, Clear2Pay
the ever-changing demands of the customer. Payments plat-
reduce process duplication, reducing in turn the need for dupli-
forms that are in existence today were just not built to cater
cate infrastructure and ultimately leading to lower cost.
for this model. The second part of the reflection centers on how to establish To tackle the issues of lowering cost, improving service, and
the roll-out of a payment hub while respecting legacy pay-
supporting new products, banks need to embark on a technol-
ment systems and are still operating to the full satisfaction of
ogy refresh. This technology refresh cannot simply be a
all concerned. Most of these are typically connections to local
replacement. The need for optimization of business processes
domestic and international payment networks. A phased
is as big, if not bigger, than the deployment of systems that
approach moving away from the siloed payment processes is
make wide use of the Internet and associated technologies.
an appropriate route, as it significantly reduces the risk of
The answer is certainly not to take the ‘big-bang’ approach. It
failure in building the next generation payment processing
is not practical to believe that there is a magical off-the-shelf
platform. Phasing involves establishing an inventory of the
solution to this problem. Though payment systems seem to
systems in place and mapping them against the new services
cater for the same needs the world over, the differences from
that need to be provided, while maintaining the end goal of a
one bank to the next are significant.
component-based payment infrastructure.
When considering this problem, it is easy to lose sight of one
The steps of establishing the phasing for the roll-out are: Map
of the most fundamental issues, the business process. Whilst
the existing bank payment systems against the optimized
all the silos that have been developed are to cater for different
component-based payment infrastructure, prioritize by identi-
types of payments, when broken down to the lowest common
fying urgent needs and selecting key components against the
denominator, a payment, irrespective of type, involves the
optimized infrastructure, and implement the new services and
debiting of one account with an amount and the crediting of
initiate gradual migration to the new infrastructure of the
another account with that same amount. We need to bring
legacy systems.
payments together by going back to the basics. Over time the payment type independence, the component
There is a way
re-use, and the volume growth will be achieved, ultimately
A bank should take a corporate view of payments as an enter-
leading to efficient straight-through-processing, single view of
prise-wide issue. They should look for a solution that abstracts
customer, integrated liquidity and risk management, all-
the payment process from the channels and that optimizes the
encompassing fraud detection, and global payment manage-
business processes irrespective of payment type, recognizing
ment solution. All this with a low cost approach of implement-
that whether a payment is a low-value batch ACH transaction
ing the payment hub infrastructure, because a bank can grow
or a high-value cross-border payment, the core processes are
its payment center in a value-based and affordable manner
mostly similar. Common processes are, for example, balance
over time.
and limit checking, account number validation, deadline and cut-off verification, branch identifier or sort code. These pro
28 - The
The solution
cesses are implemented in the main process flows irrespective
A bank payment hub is the most practical and future-proof
of payment type. It becomes important to understand the dif-
system to match a bank’s objectives and roll out approach. It
ferences, isolate those differences, and process them accord-
is a component-based payments architecture compliant with
ingly outside the main process flows in business processes that
the latest technology standards. The different elements of the
handle those unique exceptions. This is the route to radically
architecture are business process manager, business services,
Journal of financial transformation
Over the past 30 years the payments industry has grown
Whilst the world of bank payments was evolving from low
increasingly more complex, creating unwieldy silos within
value through to high value and cross-border, a totally differ-
banks. In the developed world Automated Clearing Houses
ent industry and infrastructure was forming in parallel, cards.
(ACH) were created to process low-value payments that are
Initially card associations were limited to processing credit
entered through a variety of channels, including checks and
transactions, but that has been extended to include debit
other paper instruments, telephone, file transfer, branch,
transactions attached to a debit account. Within a financial
Internet, ATM and so forth. Often these channels have their
institution, it is typical for the cards division to be completely
own infrastructure to process these payments correctly
separate from any of the other payments divisions. This led to
before they are forwarded to the Clearing House for settle-
the creation of other separate infrastructures to handle pay-
ment. Furthermore, these ACH payments vary from country to
ments via credit and debit cards, using other standards for
country and use different proprietary formats.
processing of the card payment transactions. When one considers this picture it is not hard to understand that although
The introduction of high-value payments systems, called Real
banks generate 35% of all revenues through payments, they
Time Gross Settlement, made it possible for payments to be
also incur around 40% of their total cost, according to the
cleared in real-time or near real-time. Although they have
Boston Consulting Group.
significant overlap with ACH payments, with regard to the processes that are applied, it has unfortunately not prevented
The challenge
banks and Clearing Houses from creating another whole new
The end-result of this evolution is that payments systems are
set of infrastructures, to handle, for all intents and purposes,
siloed, resulting in a huge amount of infrastructure duplica-
the same payment. As with the ACH payments, multiple chan-
tion, both hardware and software, and becoming a severe
nels are used to enter RTGS payments into the system.
maintenance headache. Systems are dated, difficult to support, and in many cases running on technology platforms that
Both ACH and RTGS are domestic payment systems and were
are unsupported or will be in the near future. Infrastructure
not created to handle cross-border transactions. In response,
duplication has not been limited to technology, but also
a number of the banks created SWIFT, The Society for World
involves staff employed to ensure smooth processing of non-
International Funds Transfer, which is a secure messaging
STP payments and to monitor and maintain the infrastructure.
infrastructure that works on the principle of correspondents. A SWIFT member bank offers access to local clearing for its
These issues translate into a variety of challenges for banks,
overseas correspondent banks, and vice versa. Once again,
such as making changes to payment systems quickly, provid-
although in essence a cross-border payment enables an
ing customers with information about the payment lifecycle,
account to be debited and an another to be credited, albeit,
to process payment related information such as purchase
for accounts that are held in different countries and different
orders, invoices, and remittance advices, to have an integrated
currencies, a completely separate infrastructure has been cre-
view of the customer, and being able to track payments end-
ated within banks to handle cross-border payments. As with
to-end. Silos require similar changes to be made in more than
ACH and RTGS payments, there are a variety of channels
one place and increase the cost of ownership.
available to the customer to initiate cross-border payments, which once again duplicated more infrastructures. In addition,
In today’s world of frantic competition, banks need the flexibil-
due to the nature and complexity of the payment instructions
ity to be able to offer new competitive products to their cus-
often a large back office staff was needed to manually enter
tomers quickly. More and more, this requires the banks tech-
and repair payments.
nology platforms to be configurable, flexible, and adaptable to
29
dations that occur prior to a payment being passed to the
ages all payments transactions originating from all of the
payment hub and passes them to the BPM with flags that
bank’s e-channels and branches. By adopting a system that is
indicate what rules have been applied to the payment. This
services as well as components-based, they migrate over
allows the BPM to determine the workflow entry point and
time and at their own pace to a future-proof platform that is
business services that need to be applied to the payment in
the center of all payment flows independent of where the dif-
order to complete the processing of the payment.
ferent components operate or who wrote or developed them. Abstracting and rationalizing the business process from the
By using this approach of CAAs, it is easier to migrate existing
channels yields immediate benefits that improve operational
functionality from existing channels into the payment hub.
efficiency. The channel aware processes, both adapters and
When first implementing a payment hub, it is likely that exist-
submitters, allow the bank to shield the underlying processes
ing channels already perform some of the required payments
from the channels and to become flexible in accepting new
processing.
channels. Future developments are not only made at their own pace but are also vendor independent, because the solu-
In this scenario the payment hub either does not need to pro-
tion allows the bank to develop their changes and new com-
vide all payment processing or the workflow can be configured
ponents in-house. Thus making it easier to react to new
such that the processing already performed in the channels is
developing business needs that present a favorable business
not performed again in the payment hub. In the longer term,
case.
it is relatively easy to migrate the functionality present in the channel into the payment hub. The functionality in the channel can be switched off and the workflow in the payment hub can be extended to include the processing previously carried out in the channel.
Channel aware submitter Channel aware submitters (CAS) are similar to CAAs. The difference being that they submit payments from the payment hub to external systems for processing. These external systems range from other bank systems (such as the host), truly external systems (such as the local ACH or RTGS system), the SWIFT gateways, and the customer delivery channels. The submitters are responsible for mapping the payments into a format expected by the target system and they also have to be able to add or delete information that is expected by the delivery channel. A good example of this is any security handshake that would need to occur for authentication purposes between two external systems.
Conclusion Using a core payments engine, a bank achieves its objective to operate an intelligent central transaction hub that man-
30 - The
Journal of financial transformation
data access layer, channel aware adapter, and channel aware
Data access layer
submitter. Standard components must be delivered together
The data access layer (DAL) is the logical infrastructure layer
with a toolkit that a bank can use to extend the functionality.
that the payment hub sits on. The business services access the data through the DAL, which describes the data elements
Business process manager
that reside within one or more physical databases. The DAL
Abstracting the workflow and business process from the
uses data resident within the payment hub and accesses infor-
underlying business services, offers the necessary flexibility
mation retrieved from sources outside. There can be data
required to roll out a component-based payment architecture.
elements that are not directly payment-related that neverthe-
The business process manager (BPM) is at the heart of the
less need to be accessible in order to provide for end-to-end
payment hub and performs the key roles of workflow and busi-
payments processing.
ness process management. The BPM has no processing capability. When a payment is received by the payment hub, the
If part of the business process is to check the limits and the
BPM identifies the type of payment and calls the appropriate
account information is held on a host, the business service
workflow and related business processes.
invoked by the BPM will retrieve the account limit for the specified account number. The business service expects a
If a new business process step needs to be added to the work-
standard logical data model, part of the payment hub that
flow, for example catering for a new regulation, it is done at
defines where limits data is retrieved from. The DAL services
the BPM layer without any change to the existing business
the different business services with the data they require, and
services. Likewise, if a business service written in an old tech-
performs the actions required to get information from the
nology needs to be replaced, this is done so without any
outside world itself.
changes to the BPM. Having this capability insulates the payment system architecture from future changes that cannot be
As with the BPM layer, the DAL needs to be abstracted from
planned for at the design and development stage, truly offer-
its physical implementation in order to provide insulation from
ing the often called for time-to-market and flexibility benefits.
future changes and from diverse implementation environments. Inevitably there are changes to the physical data loca-
Business services
tion and it is important that they do not necessarily have an
Business services are discrete components that execute units
effect on the BPM and business services. Similarly it is neces-
of work. The components are services to be used when pro-
sary to allow changes to the BPM and business services that
cessing many different payment types via different channels.
have no effect on the logical or physical deployment of the
The BPM invokes components or processes to perform certain
DAL.
tasks in the order determined by the BPM. These components are contextless, which means that they have no knowledge of
Channel aware adapters
the workflow or process. They receive a request with associ-
The channel aware adapters (CAA) manage on the one hand
ated parameters, perform their task, and return the result to
the specifics of the channel, which range from a web-based
the BPM. Provided that the existing bank applications adhere
channel, an IVR, message-based real-time payments or large
to the interface layer defined by the BPM, it is possible to re-
file-based batch payments, basically across all the different
use existing components that are already utilized within the
communication and transport protocols that exist in today’s
bank’s environment. Similarly, it is possible to use existing
fragmented payments world.
third-party components if they support the interface of the BPM or if they can be wrapped to support it.
On the other hand, the CAA manages the pre-processing vali-
31
Emerging models
Payment systems and regulation
Hans Geiger Professor, The Swiss Banking Institute, University of Zurich
Fritz Klein Independent Consultant, and Chairman, CLS Holdings/CLS Bank International
Abstract Payments and payment systems are an important function of banks and the banking system. Financial intermediation, the core role of the banking industry, can be divided into capital and payment intermediation. A lot has been analyzed, written, and said about capital intermediation, its balance sheet related risk management and regulation. Risk management and regulation related to payment systems is a newer topic. This article points out sources of risk in payments and payment systems and analyses the approach to regulation.
33
Payment systems and regulation
The term financial intermediation is generally used to describe
Because a multitude of banks exist in today’s word, all non-
the financial service industry’s role as a middleman between
cash payments for transactions, where the buyer and the
the ultimate savers and borrowers. We call this function capi-
seller do not have their accounts with the same bank, trigger
tal intermediation. In addition, a second sphere of financial
secondary transactions between the banks of the payer and
intermediation exists, which can be described as payment
the payee. These secondary transactions are processed
intermediation. It stands for the financial function of the
through street-side or inter-bank systems. Today, street-side
middleman between the payer and the payee as the monetary
systems are at the very heart of effective and efficient pay-
component in a market transaction. According to Merton and
ments. There are three basic architectures for organizing
Bodie, ‘clearing and settling payments’ is one of the six core
such systems, the network, where all banks maintain bilateral
functions performed by a financial system. Although payment
account relationships with each other (a special case of the
intermediation is a less popular subject than capital interme-
network is the correspondent banking system, where some
diation, it plays an equally important role in our economy. It is
banks act as special nodes in the network and serve other
the goal of this paper to make a contribution to our under-
banks), the centralized clearing and settlement organization,
standing of the risks and rewards of payment intermediation,
which serves as a hub for all the banks (the clearinghouse
and to discuss whether it should be regulated and supervised,
collects the payment transactions of all the banks over a
and if so, how this should be done.
certain period of time and calculates a net position for every bank. Upon settlement the clearinghouse debits or credits
Payment intermediation is highly important for the economic
the participating banks with the net amount of all transac-
wealth of the world. Without the use of effective payment
tions), and the real-time gross settlement system, an alterna-
mechanisms, today’s division of labor through market trans-
tive central solution, where the payment transactions are
actions would be impossible, Adam Smith’s invisible hand
continuously charged and credited to the accounts that the
could not perform its vital role, and the world would remain
participating banks maintain with the central system. The two
stuck in a prehistoric, self-supplying economy. In this context
systems with a centralized architecture are often operated or
it is interesting to look back on the origin of banking in Europe.
closely supervised by a central bank and serve as a transmis-
The driving force behind the development of the financial
sion mechanism for monetary policy. The centralized systems
system of pre-industrial Europe was not lending per se, but
have traditionally been organized by each individual country
payments [Kohn (2001)]. In the medieval economy, most trade
and by currency, which made international payments ineffi-
took place within local communities of people who knew one
cient, costly, and risk-prone. International payments still
another well, and such trade was largely conducted on the
depend largely on the decentralized correspondent banking
basis of trust. The subsequent development of trade among
system.
strangers in different locations introduced a new challenging
34 - The
task, which was to overcome anonymity and the differences in
The enormous innovations in information and telecommunica-
time and distance. This task required new methods of pay-
tion technologies over the last few decades led to the collapse
ment that did not depend on spatial proximity and mutual
of time and distance, the two physical dimensions that are at
acquaintance between the payer and the payee. The two prin-
the historic roots of the banking system. Yet, the collapse of
cipal financial innovations of pre-industrial Europe, the depos-
distance does not mean that geography does not matter any
it bank and the bill of exchange, evolved to address this need
more. Besides distance, geography has many dimensions that
and moved the banks into their present intermediary role,
affect trade and payments, such as laws, customs, language,
substituting the counterparty in the payment process both for
anonymity or familiarity, education, and history or path
the seller and the buyer [McAndrews and Roberds (1999)].
dependence. Technological innovations will result in a further
Journal of financial transformation
Payment systems and regulation
reduction of transaction costs and may even include the
Market risk — The risk that market prices change until the
emergence of a new type of money, digital money. Such devel-
business transaction is fully concluded. This is primarily the
opments could challenge the very role of banks as payment
result of the business practice that a transaction is concluded
intermediaries.
on day A (‘trade day’ T) and settled (fulfilled) on day B, sev-
Payments and their regulation are thus not only important for
settlement will really take place. Such a gap quite naturally
the economy as a whole, but also for the future of banks, both
exists in forward transactions, but there is also a gap of two or
in a global and local context. Payment services are one area
three days in normal cash trading [usually two days (x = 2) in
where banks still enjoy significant advantages over other com-
foreign exchange spot dealing and three days in cash securi-
petitors, not least because of their access to the street-side
ties dealing]. This type of market risk does not constitute part
systems and the central bank. Since banks are challenged at
of the payment phase, but rather of the preceding dealing
eral days later (T+x). Of course, there is no guarantee that
many fronts in capital intermediation, their role in payment
phase. It can be reduced by introducing a Central Counterparty
systems might become an important factor for future success
(CCP) into the business process or by eliminating the time gap
or failure. It would be dangerous to assume that the past
between dealing and payment. The CCP with its risk mitigating
advantages of banks in payment services will survive naturally
structure will significantly reduce the risk that settlement will
and remain unchallenged [Llewellyn (1999)]. Banks, both as
not subsequently occur. The reduction of the time gap reduces
individual firms and as an industry group, must, therefore,
the probability and the extent of market price changes.
clarify and possibly redefine their role and strategy in the area of payment intermediation. Such a strategy must determine
Credit risk — The risk that one party fulfills its side of the busi-
the role of cooperation and competition within the industry
ness transaction (delivers) and the other party does not (coun-
and towards the customers and new competitors. In doing so,
terparty or settlement risk). Counterparty or settlement risk
it will heavily influence the role of regulation and self-regula-
can be eliminated by introducing delivery versus payment, a
tion in the area of payment intermediation.
payment process where both sides of fulfillment are conducted simultaneously. If one side does not deliver, the other does
Similarly the governments, central banks, and banking super-
not pay and vice versa.
visors must make sure that regulation or deregulation promotes the workings of the invisible hand by further reducing
Liquidity risk — If one party fulfills on time and the other only
the costs and risks of exchange transactions, while at the
with delay, then liquidity risk arises in the form of the cost of
same time maintaining and enhancing the soundness and
financing the resulting gap. For the payment system as a
stability of the payment systems.
whole the liquidity risk is paramount.
Risk and risk management in payment systems
Operational risk — During the whole dealing and payment
The risks of modern payment systems stem from the chal-
process a variety of other risks exist which we summarize
lenge that trade between strangers in different locations
under the term ‘operational risk‘. This includes human, techni-
introduced into the world of trade. Risk analysis in payments
cal, process, as well as legal risk.
and payment systems must go along the line of the trading and payment process. In general, the process can be simplified
Reputational risk — Problems in the payment function can
as follows. The payment process is initiated with the underly-
also be the source of reputational risk for a bank, which
ing dealing phase being concluded. From this point onwards
manifests itself in a reduction of future growth and earnings
five types of risk arise.
potentials.
35
Payment systems and regulation
The risks in the later stages of the dealing process are related
ness transaction in the first place. Figure 1 details the different
directly to payment systems, their structure and their design.
risks and the practical possibilities to manage them.
The payment process can only be regarded as fully concluded once the receipt of the payment is reconciled. This is the
36 - The
In addition, in payment systems the phenomenon of systemic
point-in-time where the recipient has full knowledge of the
risk is important. It results from the fact that any of the above
payment received. In managing all the risks involved, there
risks can spread across a payment system from one partici-
are the three basic options, eliminating the risk, mitigating it,
pant to another, infecting participants of the payment system
or passing it on to another party. The general operational
that were not conducting business with the original source of
nature of risks in payments is underlined by the fact that there
a problem. The structures and processes of the street-side
is an option of eliminating risk. This distinguishes these risks
systems are decisive for the probability and the consequences
from other business related risks, such as loan granting,
of systemic failures that exhibit similarities with the spread of
where elimination is only possible by not conducting the busi-
diseases through contagion in biological systems.
Management action
Elimination Mitigation
Pass-on
Credit risk Risk of one-sided payment Risk of revocation of payment received Risk of a participant not meeting his overall payment obligation at the end of a settlement cycle Risk of loss of balance created by payment received Country risk / transfer risk
• Delivery versus payment • Legal concept of finality • Real-time settlement • Settlement in central bank money • N/a
• Settlement limits • Settlement monitoring • Fast reconciliation
• Payment after confirmed receipt • Pay late
• More frequent settlement cycles • Setting of participant limits • Meeting the ‘Lamfalussy principles’ for net payment systems1 • Nostro limits • Nostro monitoring • Country limits • Country monitoring
• Careful selection of direct membership in payment system
Liquidity risk Risk of insufficient liquidity (e.g. due to unexpected payment outflows) Risk of not receiving incoming payments planned for Risk of strain of liquidity by large settlement values Risk of strain on liquidity by large payments Risk of deadlock (blocking of cross-directional payments due to lack of liquidity in gross payment systems)
• Splitting of payments
• Fast access to liquidity (i.e. intraday repo procedures) • Maintenance of reserve collateral • Monitoring of payment flow • Real-time information • Fast reconciliation • Netting concepts • Circles processing (simultaneous settlement of compensating payment flows)
Legal risk Risk of revocation of payments received Risk of conflicts of law (netting) Risk of conflicts of law (cross-border) Risk of contracts not being upheld in court
• Legal concept of finality • Gross settlement
• Careful wording of contracts • External legal opinions • Careful wording of contracts • External legal opinions
Journal of financial transformation
• Inter-market swaps
1 Report of the Committee on Interbank Netting Schemes of the central banks of the Group of Ten countries, BIS 1990.
Payment systems and regulation
Management action
Elimination Mitigation
Risk of conflict with competition laws2 Risk of conflict with consumer protection laws3
• Elimination of volume discounts • Open access criteria • Vetting of contractual agreements • Vetting of contractual agreements • Transparent product descriptions • Simple and transparent price structure
Operational risk4 Risk of system breakdown Risk of loss of data centre or operational centre (e.g. terrorism, epidemic) Risk of loss of telecommunication links Risk of corrupted data and/or software Risk of unforeseen events Risk of fraud (external)
• Dual design of all technical components • Business continuity procedures • Regular rehearsals • Dual sites for IT & operations • Inter-bank alarm organization • Business continuity procedures • Regular rehearsals • Dual telecom lines from different providers • Replacement communication processes • Regular rehearsals • Independent replacement software with reduced functionality • Permanent technical control procedures • Software development procedures • Testing • Operational procedures for exception handling • Inter-bank alarm organization • Regular rehearsals • Technical message authentication • Use of strong encryption techniques • Authorization checks • Measures against counterfeiting • Vigorous prosecution (including the
Risk of fraud (internal) Risk of clerical errors
international level) • Provision of 7x24 hotlines to customers • Customer awareness campaigns • Four/six eye procedures • Vigorous prosecution • Technical plausibility checks • Four/six eye procedures
Compliance/reputational risk Risk of processing payments with a criminal background Risk of breach of confidentiality Risk of conflict with competition laws Risk of conflict with consumer protection laws Risk of negative press reports
• Know your client procedures • Technical monitoring of payments (e.g. for trigger words) • Data encryption • Definition of rules for handling of data • Contractual pass-on to clients • Elimination of volume discounts • Open access criteria • Vetting of contractual agreements • Vetting of contractual agreements • Transparent product descriptions • Simple and transparent price structure • Careful PR
Pass-on
• Contractual pass-on to clients
• Contractual pass-on to clients
• Contractual pass-on to clients
• Contractual pass-on to clients
• Contractual pass-on to clients
• Insurance • Contractual pass-on to clients
• Insurance • Insurance
• n/a
• n/a
• n/a
• n/a
• n/a
Figure 1: Risk and risk management in payments and payment systems
2 Payment systems require agreements between financial institutions that compete with each other for customers in the payment market. These agreements may be in conflict with competition law. 3 This may especially be the case combined with conflicts with competition law or when risks are contractually passed on to clients.
4 Operational risks usually exist both on the level of the street-side system and of critically important participants of the payment system (the interruption of a critically important participant may bring the entire system to a halt).
37
Payment systems and regulation
Why regulation?
It is obvious that prudential regulation and financial stability
Banks are regulated and supervised in practically every cor-
are still two very different spheres with little connections
ner of the world, although there is no consensus in academia
between the two, other than that the second serves as an
on why and how. The general argument on an abstract level
argument for the first. The fact that credit and liquidity risks
is that regulation should compensate for the market failure
inherent in payment and settlement systems have the poten-
that would prevail in a situation without regulation and super-
tial to contribute to systemic problems makes the third
vision. Market forces alone will not necessarily achieve the
motive for regulating banks so important for our subject.
objective of efficiency and safety sufficiently, because a participant does not necessarily bear all the costs and risks or
How to regulate
there is a time lag between the business opportunity (profit)
Payments can be regulated on the three levels, payers and
and the risk (cost) arising [BIS (2001)]. The other side of the
payees, banks, and inter-bank systems. Regulation is by its
coin is the risk of government failure. Regulation of economic
very nature a national task, although the international or
activities always faces the challenge of balancing between
cross-border dimension has become a major concern with
the two extremes of market or government failure, or in the
respect to effectiveness and systemic risk of payment sys-
language of the new comparative economics between disor-
tems. On the first level we look at the two original parties of a
der and dictatorship [Djankov et al. (2003)].
market transaction, i.e. the buyer and the seller, or the payer
The three specific motivations for regulating banks are the
persons in a jurisdiction. These rules are also relevant for the
protection of individuals, primarily the banks’ customers,
actors on the second and third level, discussed below. The
and the payee. Such regulation affects all natural and legal
guaranteeing the functioning of the system, and protection
most important legal frameworks are property rights, the
against systemic risks. The notion of systemic risk stands for
criminal code, the commercial and bankruptcy law. These
the danger that problems in a single financial institution
basic legal frameworks are not specific to payment systems,
might spread and that, in extreme situations, such contagion
but they are most important for the functioning of an
could disrupt the normal functioning of the entire financial
exchange economy, which is based on market transactions
system. In spite of the fact that systemic risk and contagion
and payments for such transactions. Examples of activities
are such important motives for regulation, it is striking how
that are regulated in the criminal code and that are crucial for
little this is reflected in regulatory systems. Regulation and
payments are fraud and money laundering.
supervision are still primarily directed at avoiding the failure
38 - The
of individual banks. An illustration of this fact is the new
Important aspects of payments that are regulated in the com-
Basel proposal on capital adequacy, Basel II, which states as
mercial law are the definition of time, place, identification, and
its primary goal to strengthen the soundness and stability of
the means of payment that are essential for a payment trans-
the banking system. The Basel proposal tries to achieve this
action. Where has the transaction to be completed, at the
by exclusively regulating one aspect of individual banks:
location of the payer or the payee? Is the payment legally
Solvency in the form of capital requirements. Liquidity is not
completed at the time the payer has transferred the value or
even mentioned in the Basel proposal. Using the analogy of
at the time the payment has reached the payee? Has the pay-
the health system, solvency can be viewed as the physical
ment obligation been discharged if the amount is credited to
fitness of the people within a population, whereas liquidity
the account at the payee’s bank? An aspect that has become
plays the role of vaccination. To protect a society against
decisive for the realization of the Continuous Linked
the risk of a disease, vaccination is more important than fit-
Settlement system for the risk-free settlement of forex trans-
ness.
actions is the question of finality and conditionality, which is
Journal of financial transformation
Payment systems and regulation
influenced by bankruptcy regulations in all countries involved
legal reason for supervision. In Switzerland, as in many other
in a payment transaction. All these regulatory frameworks on
European countries, the public postal service is the market
the first level are supposed to be clearly defined and well
leader in retail payments, but it is not subject to supervision
established in the legal system of developed countries. They
by the banking authorities. In general, the regulation of pay-
are not always working well in developing countries, and con-
ments can be seen as a by-product of general banking regula-
flicts still arise in cross-border payments between countries
tion. Some exceptions to this can be found in the European
with different legal systems and commercial practices. The
Union. One was the early adoption of a common position by
decree of the finality directive by the European Union in 1998
the E.U. for the introduction of a separate prudential supervisory regime for electronic money institutions in 19996.
was an important step towards an open payment area within the E.U.5 Internationally, the establishment of the United
Another area is the regulation on cross-border payments in
Nations Commission on International Trade Law’s ‘UNCITRAL
the Euro-zone, which has so far not (yet) achieved a single
Model Law on International Credit Transfers’ in 1992 was an
payment area for non-cash payments in the internal market,
essential step for making first level rules compatible between
despite the existence of a common currency area. The E.U. has
different countries.
taken and is continuing to take regulatory actions to achieve this important goal because the market forces do not provide
Regulation on the second level concerns the banks in their
incentives for a fast and effective modernization of the E.U.-
role as payment intermediaries. It can take the form of public
wide payment infrastructures.
regulation or ruling by the government or a supervisory authority, or alternatively the form of self-regulation by the
Capital adequacy rules are the first pillar of prudential super-
banking community. There are a number of reasons for banks
vision, and the regulation of liquidity plays a secondary role.
to regulate their activities themselves. One is to avoid public
The liquidity aspect of regulation is more closely linked to the
regulation. Other important incentives for self-regulation in
role of the central bank than to the prudential regulator. In
the area of payments result from the fact that payment sys-
many countries the central bank has not only the control over
tems exhibit strong network externalities. The need to process
money, but also the ultimate authority over access of banks,
and channel payments efficiently through different banks and
and other intermediaries, to central bank money and to the
the street-side system requires technical and behavioral stan-
national inter-bank system. The provisions in the new Swiss
dards, both for transactions between different banks and
National Bank Law that mandates the central bank ‘to provide
between banks and their customers. Self-regulation, and pub-
for the liquidity of the Swiss Franc money market, to supply
lic regulation, always possesses a certain risk of monopolistic
the economy with cash, to facilitate and secure the function-
impacts. Thus antitrust legislation is one important aspect of
ing of the payment system, and to contribute to the stability
regulation on the second level. At the same time, contracts
of the financial system’ are an example for such an extended role.7
concluded by banks with their customers may contain aspects of abuse of power, thus leading to demands for consumer protection.
The third level of regulation, relating to the inter-bank or street-side systems, covers the most important aspect of
Today, the regulation of payment systems is predominantly
financial stability related to payments. In order to be effective,
based on the prudential regulation of the banks, although the
regulation and supervision on the third level must cover both
arguments for banking regulation and supervision are not
national and international aspects, because contagion does
primarily oriented toward the payment function. Contrary to
not stop at national borders. In 1990 the central banks of the
deposit taking, the payment function does not constitute a
group of ten, G10, countries created the Committee on
5 Directive 98/26/EC on Settlement Finality in Payment and Securities Settlement Systems. 6 Common Position (EC) No 8/2000, adopted by the Council on 29 November 1999. Official Journal of the European Communities, 28.1.2000, p. C 26/1.
7 NBG of Oct. 2 2003, Art. 5.
39
Payment systems and regulation
Payment and Settlement Systems (CPSS) with the aim of ana-
demands objective and publicly disclosed access criteria, to
lyzing developments in domestic and cross-border payments,
secure a competitive market for the participating banks. It is
and settlement and clearing systems. In its analysis, the
also evident that the CPSS recommendations favor payment
Committee pays particular attention to financial stability, and
systems that process debit and credit transfers electronically
thereby to the linkages between institutions, infrastructures,
rather than manually and on paper. For systems using paper-
and markets. The Committee has defined seven areas of work,
based instruments, such as cheques, there are difficulties
among them the core principles for systemically important
involved in satisfying some of the core principles.
payment systems, the use of central bank money in payment systems, and retail payment instruments and systems. With
It can be concluded that the Committee’s recommendations
regard to financial stability, the development of the ‘core prin-
are highly dependent on recent technological innovations.
ciples’ is the most important activity of the Committee. The
RTGS and electronic payment transfers were not yet available
work on the principles started in 1998 and was completed in
just two decades ago. Technology is at the very heart of prog-
January 2001 [BIS (2001)]. The core principles do not consti-
ress in payment systems. Understanding and using new tech-
tute legally binding regulation; they are intended for use as
nologies properly seems more important for the efficiency
guidelines to encourage the design and operation of safe and
and stability of payment systems than the formalities of
efficient payment systems worldwide. They define the key
whether oversight is implemented by legislation, by regula-
roles and the responsibilities of central banks in applying the
tors, by the authority of central banks, or by self-regulation.
principles. The principles have a similar role as the Basel
Mastering the risks resulting from the differences in time and
framework on International Convergence of Capital Measure
space requires the skilled use of new technologies, overcom-
ment and Capital Standards for prudential supervision. And
ing the risks of anonymity and building a solid foundation of
the central banks are the orchestrators who will promote their
confidence and trust requires standards, rules, and authority
implementation.
that warrant the proper use of that technology by the participants.
The central banks should ensure that their own systems comply with the core principles and they should have the ability to
References
oversee compliance of those payment systems, which they do
• Bank for International Settlements, CPSS (2001), Core Principles for Systematically Important Payment Systems, Basel. • Djankov, S., E. L. Glaeser, R. La Porta, F. Lopez-de-Silane, and A. Shleifer, 2003, The New Comparative Economics, NBER Working Paper Series, Working Paper 9608. • Kohn, M., 2001, Payments and the Development of Finance in Pre-Industrial Europe, Hanover, NH. • Llewellyn, D.T., 1999, The New Economics of Banking, Amsterdam: SUERF Société Universitaire Européenne de Recherches Financières. • McAndrews, J., and W. Roberds, 1999, Payment Intermediation and the Origins of Banking, Federal Reserve Banks of New York and Atlanta.
not operate directly. It is also the responsibility of the central banks to cooperate with one another in promoting the safety and efficiency of payment systems. The ten principles state, among other points, that the assets used for settlement should preferably be a claim on the central bank, which facilitates the role of central bankers to fulfill the expectations. Another principle states that payment systems should have clearly defined procedures for the management of the two most important risks, credit and liquidity. Core principle number IV requires finality on the day of value, preferably even during the day. This is a clear indication that the CPSS prefers the model ‘Real-Time Gross Settlement’ (RTGS) with continuous settlement to the ‘Clearing House’ model with multilateral netting and funding at the end of the day. Principle number IX
40 - The
Journal of financial transformation
Emerging models
Why has stored-value not caught on?
Sujit Chakravorti1 Senior Economist, Federal Reserve Bank of Chicago
Abstract Why have general-purpose stored-value cards been unsuccessful in penetrating the U.S. market? Three necessary conditions for a payment instrument to be successful are discussed: consumers and merchants need to be convinced of its advantages over existing payment alternatives for at least some types of transactions; payment providers must convince consumers and merchants simultaneously of its benefits to achieve critical mass; and assure them that adequate safety and security measures have been implemented. This article discusses the credit card industry’s success in meeting these necessary conditions and general-purpose stored-value issuers’ failure to meet them to date.
1 I thank Victor Lubasi and Leonardo Mayer for their excellent research assistance. The views expressed are the author’s and should not be attributed to the Federal Reserve Bank of Chicago or the Federal Reserve System.
41
Why has stored-value not caught on?
we pay. These four major innovations are: coins (4,000 years
thorized user is able to use the payment system for financial
ago), checks (800 years ago), paper money (more than 100
gains or a participant in the payment process presenting a
years ago), and the payment card (over 50 years ago).
monetary claim that is not backed by the value stated.
Three important conditions must be met before stored value is
An important issue with credit and fraud risk is the allocation
widely used. Firstly, consumers and merchants need to be con-
of monetary losses when it occurs. Consumers and merchants
vinced that stored value is superior to existing payment instru-
prefer that the liability lies with the payment service provider.
ments for certain types of payments. Generally, stored value
Payment instruments with this characteristic may also pene-
payments substitute for cash payments. Stored value has been
trate the market quicker. In the case of credit cards, govern-
a successful alternative to cash in closed loop systems, such as
ment regulations determine the maximum liability to the
transportation service purchases, coffee purchases at certain
consumer if the card is used by an unauthorized user. Today,
popular coffee chains, and for purchases on university cam-
the card networks have further reduced the consumer’s liabil-
puses and military bases. Nilson (2003) estimates that prepaid
ity for unauthorized use to zero.
cards, which include stored-value cards for limited use and phone cards, accounted for U.S.$55 billion in 2003 and pre-
Is it better?
dicts that this figure will grow to U.S.$146 billion in 2007.
For new payment products to succeed, they need to provide
Secondly, as with the introduction of any new payment instru-
time being profitable for payment providers in the long run.3
ment, to achieve critical mass, consumers and merchants
Issuers of new payment instruments usually target a segment
need to be convinced simultaneously. That is, consumers will
of the payment services market where their product is supe-
benefits to both consumers and merchants, while at the same
not use stored value unless a sufficient number of merchants
rior to existing alternatives. A recent example of a payment
accept it and merchants will not accept it until a sufficient
product that has benefited both consumers and merchants is
number of consumers use it. An example of the inability of a
the payment product provided by PayPal.com to buyers and
payment instrument to overcome the chicken-and-egg prob-
sellers for transactions on online auctions sites such as eBay.
lem is the Susan B. Anthony dollar coin. Because coins remain
Because many relatively small merchants were not usually
in circulation much longer than bills, they are less expensive
equipped to process credit and debit card transactions and
for currency issuers to provide in the long run. Unlike most
checks were associated with high settlement risk, a payment
countries, the United States has been unsuccessful in replac-
enabler like PayPal.com was able to intermediate these trans-
ing lower denomination bills with coins. McAndrews (1997)
actions resulting in both parties being better off.
argues that Canada, like other countries, was eventually successful with its dollar coin because the central bank started to
Credit cards
withdraw notes from circulation.
In the early 1970s, some financial observers predicted that credit cards were not viable in the long run. One such observ-
42 - The
Thirdly, with any payment instrument, consumers, merchants,
er argued that credit cards were ‘a temporary but probably
and financial institutions are concerned with credit and fraud
unavoidable retreat in the campaign to develop an efficient
risk. For our purposes, credit risk is the risk that the payee is
domestic payments mechanism’ [Hester (1972)]. Today, credit
unable to convert a payment into good funds. The inability to acquire good funds may result from the payer, a payment
card transactions rank second behind checks in terms of the number of non-cash transactions in the United States.4 There
intermediary, or the issuer’s inability to process or make good
were 17.86 billion general-purpose charge and credit card
on its obligation to deliver. Fraud risk is the risk that an unau-
transactions accounting for U.S.$1.608 trillion in the United
Journal of financial transformation
2 General-purpose stored-value payment cards are defined as those that are widely adopted by merchants and consumers where the value resides on the card and is transferred to the merchant’s terminal at the time of purchase. Unless otherwise stated, stored-value cards will be short for widely-accepted stored-value cards in this article.
Why has stored-value not caught on?
Advances in computing power, electronics, and telecommuni-
financial institutions have invested significant amounts of
cations have improved the way we live. Now such advances
money into stored-value technology in an effort to provide
have started to change the way we pay. Technological
electronic substitutes for government-issued physical cash.
advancements now make it possible for consumers to pur-
MasterCard reportedly had invested over U.S.$150 million to
chase goods with electronic bits of information representing
purchase 51 percent of Mondex International, an electronic
money, commonly referred to as stored value. The value may
cash system developed in the United Kingdom by National
be stored on microchips embedded in plastic cards that look
Westminster Bank [Hansell (1998)]. National Westminster
similar to credit cards. This type of stored value device is
spent more than U.S.$100 million developing Mondex [Stouffer
called a smart card. According to an article seven years ago,
(1996)].
‘Smart cards are set to revolutionize payment systems and provide a plethora of new opportunities’ [Talmor and Timewell
Stored-value issuers hope to earn interest from outstanding
(1997)]. Another article in the popular press stated that ‘Cash
stored-value balances, earn fees from merchants, and possibly
is dirty, inefficient, and obsolete. Smart cards, digital cash and
revenues from advertising on the physical card. However, issu-
a host of electronic currencies will soon replace pocket money’
ers will have to convince consumers and merchants why they
[Gleick (1996)]. This article asks the question: Why have gen-
should use stored value. Issuers argue that their product
eral-purpose stored-value cards not been widely adopted as some analysts had expected?2 I will compare the credit card
would be more convenient for consumers and reduce costs of processing payment for merchants.
industry’s success in meeting three conditions necessary for widespread adoption with the stored value issuers’ failure in meeting these conditions to date in the United States.
Most analysts agree that the two largest U.S. stored-value trials, the Atlanta Olympic Games and the Upper West Side of Manhattan, failed in convincing consumers and merchants of
Financial analysts have predicted the death of cash and other
the benefits of using stored value over existing payment alter-
paper-based payment instruments for many years. Cash
natives. The Economist (1998, 73) concluded that, ‘Electronic
usage has started to decline. According to an American
money has thus turned out to be a solution in search of a
Bankers Association/Dove Consulting Study, cash usage has
problem.’ While general-purpose stored-value has not been
declined from 39 percent to 32 percent for in-store payments
successful in the United States, many European countries
[Sapsford (2004)]. Part of this decline results from greater
have implemented such payment instruments with varying
acceptance of payment cards, such as credit and debit cards,
degrees of success. This article will address some factors that
at merchant locations that traditionally had not accepted
have lead to adoption of stored-value payment instruments in
them. In the United States and most parts of the world, limit-
some countries but not others.
ed-use stored-value cards have been successful as cash substitutes for some niche markets, such as transportation systems, university campuses, and military bases.
The necessary conditions Consumers and merchants are reluctant to change their preferences towards payment instruments. In the context of the
Smart card issuers along with producers of the technology
issuance of new coinage, Jevons (1875) wrote, ‘No one can
have made sizeable investments to establish smart cards as a
possibly understand many social phenomena unless he con-
viable payment instrument. The migration to chip cards from
stantly bears in mind the force of habit and social conventions.
magnetic stripe ones have been aided by the reduction in the
This is strikingly true in our subject of money.’ Furthermore,
cost of producing smart cards. Payment card organizations,
Evans and Schmalensee (1999) observe that in the last 4000
such as MasterCard and Visa, along with banking and non-
years there have been only four major innovations in the way
3 Chakravorti and Kobor (2003) argue that payment products may not have to be profitable as a stand-alone product, but sufficiently add value to a bundle of banking and payment services.
4 In certain payment segments, such as in-store purchases, credit cards have surpassed checks. In other market segments, such as Internet payments, credit cards continue to be the preferred payment instrument although signature-based debit cards have started to catch up. However, for recurring bill payment and businessto-business payments checks continue to make up the lion’s share of payments [Chakravorti and McHugh (2002) and Chakravorti and Davis (2004)].
43
Why has stored-value not caught on?
Consumers are unlikely to use stored value for purchases
stored value at relatively low cost.
where they use checks, credit or debit cards because they risk losing monetary value if the stored value is lost or stolen and forgo the opportunity to earn interest on their funds before they spend them.8 Evidence from Scandinavian countries suggests that greater penetration of debit cards especially for low-value transactions has resulted in greater reluctance to use stored value products [Van Hove (2004)]. Thus, stored value may only replace a shrinking number of cash transactions. Often successful adoption of a stored-value card is associated with eliminating alternative payment methods completely. In the Dutch cities of Prumerend, Nijmegen, and Rotterdam, the use of Chipknip, a general-purpose stored value payment system, became the only way to pay for street parking. As a result, parking accounted for 31 percent of the total number of Chipknip payments [Van Hove (2004)]. Stored-value cards may also become the only payment option for some French parking lots. Merchants may benefit from a lower volume of cash transactions because they are more prone to safekeeping concerns and on average take longer to perform than stored-value ones. Lucas (1994) states that employee theft can account for up to 4 percent of cash sales for primarily cash-based transit systems. In addition, some merchants would benefit from quicker transactions because the transfer between the merchant’s stored-value machine and the consumer’s smart card would be faster than alternative payment forms [Poon and Chau (2001)]. However, merchants may be the most reluctant to use storedvalue technology. Merchants may face large transition costs in acquiring the necessary hardware to accept stored value and training their staff. Some analysts argue that the initial investment may be relatively small compared to the potential cash savings, especially since acceptance terminals for other payment cards would need to be replaced over time. When these terminals are replaced, they could be fitted for acceptance of
44
5 Chakravorti and Emmons (2003), Chakravorti and Shah (2003), and Chakravorti and To (1999) discuss the incentives for consumers to use credit cards and merchants to accept them.
6 In addition to security concerns, foreign travelers may need to convert their cash into local currency prior to making purchases and convert the remaining foreign currency back to home currency at the end of their trips. However, some types of transactions, especially small ones, may require cash for payment. 7 For a discussion of payment instrument cost, see Food Marketing Institute (2000).
Why has stored-value not caught on?
States in 2002 [Nilson (2003)].
Stored-value Limited-use stored value systems have been successfully
General-purpose charge and credit cards have existed for over
adopted for transportation systems, such as the Bay Area
40 years. Unlike charge cards, credit cards allowed consumers
Rapid Transit system in the San Francisco Bay Area and the
to pay their monthly charges in installments. Today, credit
Metro in the Washington D.C. area. Some transit authorities
cards benefit consumers and merchants and are profitable to payment providers.5 Credit cards serve two primary functions
have introduced smart cards that can be used as proximity payment devices. In Hong Kong, the transit authorities have
for consumers, they allow consumers to purchase goods and
introduced smart cards as the sole payment device and have
services (serves as a payment instrument) and they extend
found that more passengers can be processed in a given time
credit to consumers lacking sufficient funds even if they
span resulting in reduced queues [Poon and Chau (2001)].
choose to pay their balances in full (serves as a credit instrument). Consumers may also prefer to use their credit cards
Two notable U.S. general-purpose smart card trials were con-
because of frequent-use awards or dispute-resolution servic-
ducted during the last 8 years, Upper Westside of Manhattan
es. Furthermore, consumers can use them almost anywhere in
and the Atlanta Olympic Games. By most accounts, these tri-
the world. As a result, credit cards may be preferred to cash
als were not successful based on the usage rates and the lack
or travelers’ checks as a secure, widely accepted payment instrument for foreign travelers.6
of long-term adoption in the United States. However, some lessons can be learned. In both trials, consumers and merchants were given incentives to use the product. In the
Although the credit card is the most expensive payment
Manhattan trial, converted laundry machines accounted for
instrument to accept, merchants benefit from credit card acceptance, justifying their relatively high cost.7 For charge
ue cards [Van Hove (2001)]. In Atlanta, stored-value cards
and credit card purchases, most merchants enjoy payment
were more successful when merchants did not previously
guarantees from card issuers if they take the proper authori-
accept payment cards, such as fast-food restaurants and con-
30 percent of all transactions conducted with the stored-val-
zation steps. Merchants also benefit from greater sales and
venience stores [Bank Systems & Technology (1996)].
profits. In a survey of retailers, 83 percent thought accepting
Therefore, stored value cards are popular with consumers
credit cards increased sales and 58 percent thought their
and merchants at merchant locations that had traditionally
profits increased from accepting them [Ernst and Young
accepted only cash.
(1996)]. These greater sales are generated in part because consumers may not have sufficient cash on hand.
Similar to credit cards, for stored value to be successful, con-
Financial institutions earn revenue from the merchant dis-
all perceive a benefit from its use. A lucrative niche market
sumers and merchants along with financial institution must count, interest income from consumers who borrow beyond
segment for stored-value cards are unmanned point of sale
the payment cycle, and other fees from additional services
purchases, such as vending machines, parking meters, and
provided. However, there are risks that financial institutions
fares for transportation services. A key issue for consumers
take when issuing credit cards. Investment in new payment
in such purchases is the availability of exact change. Coca-
products may not immediately generate a positive return. In
Cola estimates that a significant portion of their potential
the case of Bank of America, fifteen months after launching
sales never took place because customers did not have the
its BankAmericard, it officially lost U.S.$8.8 million dollars. If
exact change [Clemons, Croson, and Weber (1997)]. In other
hidden costs, such as advertising and overhead, were included
cases, consumers may overpay for services, such as parking
the loss was closer to U.S.$20 million [Nocera (1994)].
and tolls.
8 For a general discussion about the preferred payment instrument by consumers and merchants, see Chakravorti (1997) and Humphrey, Pulley, and Vesala (1996). 9 For a discussion about the interrelated demand for payment services see Baxter (1983).
10 See Economides and Himmelberg (1995) for a discussion of critical mass in the context of network goods.
45
Why has stored-value not caught on?
and issued cards to applicants if they had a job. Because
Stored-value
cardholders initially did not incur any of the costs associated
To promote usage, payment instrument providers entice both
with credit card transactions, they were easily convinced to
consumers and merchants with incentives. In the two largest
use the cards.
U.S. general-purpose stored-value trials, issuers gave consumers monetary value to promote its use. At the Atlanta OlymÂpic
Card issuers had more difficulty bringing merchants on board
Games, some stored-value cards were given away with five dol-
because the merchants were charged a fraction of the pur-
lars of purchasing power, but cardholders preferred to keep
chase price. Diners Club managed to convince twelve restau-
them as souvenirs. In New York, Citibank employees handed
rant owners to accept their card at the time of launch. Bank
out cards with U.S.$5 of value to passersby. Merchants also
of America started with 300 merchants. Larger merchants
received the necessary equipment at subsidized rates and
were unwilling to pay the merchant discount. The first large
may not have paid the full merchant discount.
department store chain to accept third-party credit cards was J.C. Penney in 1979 and widespread acceptance by grocery
To increase awareness of smart card technology, some finan-
stores has only occurred recently. On the other hand, smaller
cial institutions in other countries have started to use existing
merchants that granted their customers credit were willing to
payment instruments, such as ATM and credit cards, to pig-
pay the fee to reduce their accounting, collection, and billing
gyback stored value by placing microchips on these cards. For
costs.
example, financial institutions in Belgium and Finland have started to put microchips on ATM cards. In these countries,
To expand the geographic coverage of its cards, Bank of
consumers must use stored value to pay for parking meters,
America began to license the BankAmericard through Bank
calls from public phones, and bus tickets [The Economist
America Service Corporation to out-of-state banks. Banks
(1998)]. These uses of stored value may increase consumers’
would pay a U.S.$25,000 entry fee to Bank of America and a
awareness and comfort level.
small royalty to support a national advertising campaign to become members of the network. Each bank would enlist its
Is it safe and secure?
own merchants and customers. The main goal of these licens-
The sustainability of a new payment instrument is critically
ing agreements was to increase the number of consumers
dependent on the containment of credit and fraud risk. The
using the card and the number of merchants accepting the
success of any payment system is related to the faith and
card. Bank of America benefited from BankAmericard holders
confidence that participants have in it. Payment providers
from other states making purchases from their merchants and
should convince consumers and merchants that they can con-
from their customers making purchases from merchants of
vert their claims into good funds with minimal risk. If the pay-
their licensees.
ment provider becomes bankrupt and has payment obligations outstanding, consumers and merchants may face sig-
Card issuers used innovative ways to simultaneously con-
nificant losses. To limit credit risk, some European regulators
vince consumer and merchants of the cards’ benefits. The
have argued that stored value should be only provided by
more consumers that card issuers convinced, the more mer-
regulated financial institutions.
chants were willing to accept it. Although credit cards were eventually successful in overcoming the chicken-and-egg
46 - The
Along with credit risk, payment providers are concerned with
problem, Osterberg and Thomson (1998) argue that critical
containing fraud risk. Roberds (1998) describes two major
mass was only achieved in the late eighties when its growth
forms of fraud. In the first case, the buyer presents a mone-
exploded.
tary claim that is not backed by the value stated. For example,
Journal of financial transformation
11 For an excellent discussion on why the debit card has been slow to penetrate the marketplace see Caskey and Sellon (1994). 12 Recently, around 5 million U.S. retailers settled a case against MasterCard and Visa over the tying of association credit and debit card products. As part of the settlement, the card associations can no longer tie their credit and debit card products. For more about this case see Chakravorti (2003).
Why has stored-value not caught on?
Financial institutions should also benefit from the shift to
transactions start to gain popularity.11 One type of debit card,
stored-value from cash. The migration to electronic substi-
commonly referred to as PIN-based, uses ATM networks to
tutes for cash may provide greater profit opportunities for
process transactions at the point of sale and is also an ATM
financial institutions in terms of cost reductions associated
card. Issuers were initially unsuccessful at convincing a suf-
with security and transportation. In addition, financial institu-
ficient number of merchants to participate primarily because
tions may benefit from income generated from issuing and
of the additional cost of installing card readers and the lack
distributing the stored value and the interest income from
of interoperability among the different ATM networks. Today,
outstanding stored value.
one in three merchants has point-of-sale terminals needed to process PIN-based debit cards [Nilson (2003)]. Furthermore,
Can it achieve critical mass?
consolidation of ATM networks and the introduction of
Payment instruments have two distinct sets of users, consum-
shared networks also increased the appeal of PIN-based debit
ers and merchants, that simultaneously demand payment
cards to merchants.
services. Consumers benefit more from an increase in the number of merchants that accept the payment instrument
Another debit card innovation that allowed greater market
than from an increase in the number of consumers that use it.
penetration was the introduction of the signature-based debit
Similarly, merchants benefit more from an increase in the
cards issued by the credit card associations. These debit
number of consumers that are willing to use it than the num-
cards use the existing credit card network infrastructure to
ber of merchants that accept it. In other words, the consum-
process and settle transactions. Because credit card networks
er’s and the merchant’s demand for the payment service are interrelated.9 These types of services are often called two-
were already extensive and merchants faced no new setup costs, these cards were able to penetrate the market much
sided because usage of these services is dependent on both
quicker. In addition, to promote acceptance of the signature-
sides being on board.
based debit cards, the card associations required all mer-
Payment services can be viewed as network goods. A good is
chants accepting their credit cards to accept their debit cards.12
defined as a network good if a user benefits from an increase in the number of users of that good [Farrell and Saloner
To overcome the chicken-and-egg problem, debit card provid-
(1985), and Katz and Shapiro (1985)]. For example, telephones
ers used existing technologies that were familiar to both
and fax machines are network goods because existing users
merchants and consumers. To increase consumer usage
benefit from an increase in the number of people that they
many financial institutions started to issue ATM cards that
can communicate with. Furthermore, a sufficient number of
were both PIN-based and signature-based debit cards. Thus,
users is required for the network good to survive. Economists define this sufficient number as a critical mass.10 Both credit
with the ATM customer base and the use of the existing
cards and stored value exhibit characteristics of network
cards were able to overcome the chicken-and-egg problem.
credit card network by signature-based debit cards, debit
goods.
Credit cards The problem of a network good achieving critical mass can be
Charge and credit card issuers used various techniques to
described as a chicken-and-egg one. An example of a good
overcome the chicken-and-egg problem. To achieve a critical
that required a long time to overcome the chicken-and-egg
mass of consumers, Bank of America mailed active cards to
problem is the debit card. Although the first debit card pilot
their existing customers. Not having a customer base to
was conducted in 1966, only many years later did debit card
solicit, Diners Club initially handed out leaflets door to door
47
Why has stored-value not caught on?
Stored-value
being added to identification cards or existing payment instru-
The most powerful deterrent against fraud in stored-value
ments where value can be stored and used to make purchases.
systems is the technology. Smart card technology may be
Such types of cards exist in closed settings, such as university
more secure than cash for merchants and offer issuers great-
campuses where students may use the stored-value feature to
er protection from counterfeiters than magnetic stripe tech-
make photocopies where other alternatives are not as conve-
nology. To prevent theft of coins from public phones in France,
nient. Alternatively, merchants using stored value in closed
callers were required to use smart cards. The major credit
systems, such as transportation systems, could enter into
card companies are considering smart card technology as a
agreements with other merchants to broaden the acceptance
replacement for magnetic stripes to reduce credit card fraud.
of the payment instrument.
Stored-value issuers want to limit or perhaps eliminate the
To achieve a critical number of consumers for stored value,
possibility that outsiders can replicate the underlying value
financial institutions may offer a stored-value enhancement to
and inject it into the system. One of the largest known cases
their existing debit and credit cards. Similar to the debit card,
where a stored-value system was compromised occurred in
where issuers used the existing ATM and credit card networks,
Japan involving Pachinko parlors, where the less secure mag-
by piggybacking on existing payment cards, stored value could
netic stripe technology was used. Criminal organizations were
benefit from economies of scope. While in many European
able to create stored value that they did not purchase. As a
countries, financial institutions replaced existing debit cards
result stored-value issuers are said to have lost at least
with ones with stored value capabilities, Van Hove (2004)
U.S.$600 million [Pollack (1996)].
states that many of these unsolicited stored value enhance-
Realizing that the most sophisticated technology to prevent
may be helpful in achieving a critical mass of potential users,
fraud may not be impenetrable, stored-value issuers are con-
it is clearly not a sufficient condition for widespread adoption.
ments remain largely unused. Therefore, while such a strategy
sidering other preventive measures. While online real-time verification would defeat the purpose of stored value, most
Experiences in Europe suggest that government mandates
issuers require redemption of the underlying value after each use.14 In these systems, fraud could be detected sooner than
may increase the acceptance of smart cards. However, even
in systems where stored value is redeemed less frequently.
ponent remain small as a percentage of total transactions. As
However, given the relatively small amount of monetary value
with the introduction of other payment instruments, stored
with such intervention, usage rates of the stored-value com-
transacted with stored value, there may be little incentive to
value cards will require some time before they achieve critical
commit fraud.
mass. While Van Hove (2004) argues that stored value has not to date achieved the desired market penetration, he identifies
Will stored-value succeed?
certain types of merchants as ideal candidates for stored
Given the comfort and convenience that consumers have with
value. These types of merchants have at least one of the fol-
existing payment instruments and ongoing improvements to
lowing characteristics: payments are time-critical (public
reduce the cost of accepting them, consumers and merchants
transport), there are high cash handling costs (vending
in the United States may perceive little benefit from stored
machines), or there are vandalism problems (parking meters
value as a stand alone payment instrument. Thus, unless con-
and payphones).
sumers are forced to use it by merchants, the widespread use of stored value as a stand alone point-of-sale payment instrument is unlikely in the United States. However, microchips are
48 - The
Journal of financial transformation
13 Today, NBI is known as Visa. 14 A notable exception is the Mondex system which allows consumers to exchange value among themselves without third-party intervention. For more details about Mondex see Clemons, Croson, and Weber (1997).
Why has stored-value not caught on?
in a check transaction, the consumer may write checks with
1960, Bank of America’s losses from fraud and defaults were
insufficient funds in his account. The other type of fraud
nearly U.S.$9 million or 15 percent of their volume. Fraud was
involves the buyer using a monetary claim belonging to some-
committed in various ways, including consumers using cards
one else.
to make purchases that they did not intend to pay for, cards
While credit and fraud risks are difficult, if not impossible, to
merchants sending in credit slips for nonexistent purchases.
being stolen from the mail and used to make purchases, and eliminate, adequate disclosure of which participant bears the loss is critical to the sustainability of any payment instrument.
Banks implemented several policies to limit fraudulent uses.
If payment providers cannot adequately guard against unau-
Banks required that merchants call their financial institution’s
thorized use, resulting losses may lead them to leave the
credit centers when purchases were above a certain amount,
industry and lead consumers and merchants to lose confi-
known as floor limits. Many banks provided merchants with
dence in using that type of payment instrument. Furthermore,
hot lists that identified delinquent accounts. Eventually,
if consumers and merchants perceive that they are more lia-
Congress outlawed the mailing of unsolicited credit cards by
ble for payments made with a new instrument, they may be
financial institutions in an effort to limit fraudulent use.
less willing to use it.
However, these measures were not sufficient.
Credit cards
The use of computers and telecommunications in the autho-
Historically, credit and fraud risks have been challenging for
rization process allowed credit card organizations and their
credit card issuers to contain and have led to a number of issu-
members to contain fraud. In 1972, National BankAmericard,
ers leaving the business. Technological advancements along
Inc. (NBI), the credit card organization spun off by Bank of
with government regulations significantly reduced these risks.
America, introduced a nationwide network linking computers
However, credit card networks continue to improve and intro-
via telephone lines to authorize credit card transactions at
duce new measures to mitigate these risks.
the point of sale.13 Although the system cost U.S$3 million to
Credit risks are contained by guidelines and rules at various
U.S.$30 million in the first year [Nocera (1994)]. The initial
levels in the credit card network. The risk that a financial insti-
authorization system still involved humans checking com-
tution is unable to meet its payment obligation is controlled
puter screens for the status of the customer’s account.
build and implement, it saved members of NBI at least
primarily by the card associations. Because the cost of losing
Today, the process is completely automated and most trans-
their reputation is so high, the associations impose guidelines
actions are authorized prior to purchase. Further improve-
governing the distribution of losses if a member institution is
ments to the physical card, the network, and the monitoring
unable to meet its obligations. Credit risk at the consumer and
of charges have led to significant reduction in losses from
merchant level is primarily contained by policies of the finan-
fraud.
cial institutions involved. Today, financial institutions use more rigorous methods to determine creditworthy consum-
Although credit and fraud risk have not been eliminated, suf-
ers. In addition, part of the interchange fees charged by card-
ficient steps have been taken to assure consumers and mer-
issuers to merchant banks covers the credit risk the issuer
chants that they face minimal liability when using and accept-
faces from consumers unable to pay their obligations.
ing credit cards. The adoption of system wide guidelines
Fraud was a major factor in the early years of charge and
reduced these risks in the credit card network.
along with the aid of real-time online processing has greatly credit cards. Evans and Schmalensee (1993) report that in
49
Why has stored-value not caught on?
Conclusion
References:
This article explored three necessary conditions for the viabil-
• Bank Systems & Technology, 1996, “Olympic cash card pilot results are in: Merchants the key to program’s success,” 33:9, September, 8 • Baxter, W. F., 1983, “Bank interchange of transactional paper: Legal and economic perspectives,” Journal of Law & Economics, 26, 541-588 • Caskey, J. P., and G. H. Sellon, 1994, “Is the debit card revolution finally here?” Federal Reserve Bank of Kansas City Economic Review, First Quarter, 79-95 • Caskey, J. P., and S. St. Laurent, 1994, “The Susan B. Anthony dollar and the theory of coin/note substitution,” Journal of Money, Credit, and Banking, 26:3, 495-510 • Chakravorti, S., 1997, “How do we pay?” Federal Reserve Bank of Dallas Financial Industry Issues, First Quarter • Chakravorti, S., 2003, “Theory of credit card networks: A survey of the literature,” Review of Network Economics 2:2, 50-68 • Chakravorti, S. and E. Davis, 2004, “An electronic supply chain: Will payments follow?” Federal Reserve Bank of Chicago Fed Letter, September • Chakravorti, S. and W. R. Emmons, 2003, “Who pays for credit cards?” Journal of Consumer Affairs, 37, 208-230 • Chakravorti, S. and E. Kobor, 2003, “Why invest in payment innovations?” Federal Reserve Bank of Chicago Occasional Paper Series, 1B • Chakravorti, S. and T. McHugh, 2002, “Why do we still write so many checks?” Federal Reserve Bank of Chicago Economic Perspectives, 3rd Qtr, 44-59 • Chakravorti, S. and A. Shah, 2003, “Underlying incentives in credit card networks,” The Antitrust Bulletin, Spring, 53-75 • Chakravorti, S. and T. To, 1999, “A theory of credit cards,” Federal Reserve Bank of Chicago Working Paper Series, WP-99-16 • Clemons, E. K., D. C. Croson, and B. W. Weber, 1997, “Reengineering money: The Mondex stored value card and beyond,” International Journal of Electronic Commerce, 1:2, 5-31 • Economides, N. and C. Himmelberg, 1995, “Critical mass and network size with application to the U.S. fax market,” New York University, Working Paper No. EC-95-11, August • The Economist (1998), “Keep the change,” November 21, 73-74 • Ernst & Young (1996), “Survey of retail payment systems,” Chain Store Age, January • Evans, D. S., and R. L. Schmalensee, 1993, The economics of the payment card industry (Cambridge, Mass.: National Economic Research Associates, Inc.). • Evans, D. S., and R. L. Schmalensee, 1999, Paying with plastic: The digital revolution in buying and borrowing, (Cambridge, MA: The MIT Press) • Farrell, J., and G. Saloner, 1985, “Standardization, compatibility, and innovation,” Rand Journal of Economics,16, 70-83 • Food Marketing Institute, 2000, It all adds up: An activity based cost study of retail payment instruments (Washington, DC: Food Marketing Institute) • Gleick, J., 1996, “Dead as a dollar,” New York Times Magazine, June 16, 9-16 • Hansell, S., 1998, “Got a dime? Citibank and Chase end test of electronic cash,” New York Times, November 4, Business Section, 1 and 4 • Hester, D. D., 1972, “Monetary policy in the “checkless” economy,” Journal of Finance, 27, 279-93 • Humphrey, D. B., L. B. Pulley, and J. Vesala, 1996, “Cash, paper, and electronic Payments: A cross-country Analysis,” Journal of Money, Credit, and Banking, 28:4, 914-39 • Jevons, W. S., 1875, Money and the mechanism of exchange (New York: D Appleton & Company) • Katz, M. and C. Shapiro, 1985, “Network externalities, competition and compatibility,” American Economic Review, 75:3, 424-40 • Lucas, P., 1994, “The card that came in from the cold,” Credit Card Management, 7, 40 and 42 • McAndrews, J. J., 1997, “Network issues and payment systems,” Federal Reserve Bank of Philadelphia Business Review, November/December, 15-25
ity of a new payment instrument. A new payment instrument may take longer for consumers to accept because of the complex set of interactions that occur among participants. It must provide benefits not provided by existing ones for at least certain types of transactions. Consumers and merchants must be convinced simultaneously of its benefits and may require incentives to change their behavior. Finally, the payment instrument should be relatively safe and adequate measures against credit and fraud risk should be adopted. While credit cards were successful in meeting these three necessary conditions, stored-value cards have yet to meet them. However, in markets where limited-use stored-value cards have been successful, they are generally a substitute for cash. They are popular with consumers when exact change is required. In some cases, as with the dollar coin, significant market penetration may not occur unless consumers are forced to adopt stored value, such as payment of transportation services and parking fees. They are popular with merchants when cash handling costs are high and other alternatives are not available for payment. Today, the most successful limited-use stored value operators have started to leverage their expertise to expand acceptance of their product beyond its initial use. The Octopus stored-value system in Hong Kong was expanded from payment for transportation services to include purchases at non-transit related merchants. The introduction of a new payment instrument requires sufficient time to educate consumers and merchants of the benefits of migrating from existing payment options. If stored value is to succeed, both consumers and merchants must be convinced of its benefits.
50 - The
Journal of financial transformation
Emerging models
Online payments systems for e-commerce1
Caroline Paunov Consultant, OECD
Graham Vickery Head of the Information Economy Unit, OECD
Abstract With the slow uptake of e-commerce in the business-to-con-
networks and wide user-bases. However, traditional credit
sumer (B2C) sector, the initial highly optimistic growth predic-
card payments have deficits in the areas of security and ano-
tions of e-commerce soon came to be revised downwards.
nymity. Moreover, the payment means is not ideal for auc-
With the gain in maturity of e-commerce there has been a rise
tions and micropayment markets. Alternative systems in
in suitable products for online sales and some trend towards
some countries are debit cards and payments via online bank-
increased consumer confidence. For these positive develop-
ing. Given limited user bases, electronic cash is not a feasible
ments to finally lead to the long-expected growth in market
payment option. In contrast, the mediating service PayPal has
share, things will crucially depend upon the existence of well-
been successful particularly in serving the auction market.
developed online payment solutions for e-commerce. This
Furthermore, large user bases and high accessibility render
article aims to brush some broad conclusions on this issue.
mobile devices as potentially important future means for
A first observation is that despite the existence of a variety of
Internet options, play an important role in providing micropay-
conducting payments. In particular, they may, together with payment means credit cards are the dominant system for
ment solutions. For these payments, there is still a way to go
online money transfers. This is a consequence of advanta-
to establish a clear alternative to the non-optimal subscrip-
geous characteristics, most importantly the long-established
tion model.
1 The views expressed in this article do not necessarily represent the views of the OECD or its member governments.
51
Online payments systems for e-commerce
of all analyzed websites. An important aspect to bear in mind
information online.
is the divergence of importance of payment means across countries. Figure 1 shows that for individual countries, as for instance Finland, other payment systems, notably debit cards but also e-banking facilities, are also of importance.
Advantages of credit cards The dominant market position of credit cards is easily explained by their distinguishing characteristics. The longestablished network combined with a wide user-base constitutes a very important advantage. This factor gains particular significance in the online market, because it is in its nature a network infrastructure market that provides a platform that permits money transfers between the buyer and seller. As such, the payment system’s value rises with the size of its user-base, both on the buyer and the seller sides. Further benefits arise from established presence and consumer experience with the payment means as well as established and clear perspectives on liability. A further key advantage lies in the existing international dimension of credit cards, which permit payments across national borders easily. These advantages effectively constitute the basis for a strong market position, which alternative payment methods will have to provide and compete with.
Security of credit card payments Despite their widespread use and acceptance, traditional online credit card payments have deficits with respect to the security of transactions because of weak authentication and the necessity to transfer detailed financial information for payment3. As a result, fraud rates for online credit card transactions are higher than for payments through the offline channel. Yet the losses to fraud are not currently assuming alarming numbers. However, household surveys across OECD countries constantly find that security concerns are key factors explaining consumer unwillingness to use online payment instruments. To give one illustration, PaymentOne survey data suggest that nearly three out of four U.S. consumers do not use credit card payments online because of security (Figure 2). And this is particularly related to providing credit card
52 - The
Journal of financial transformation
2 The survey was conducted by PricewaterhouseCoopers on behalf of the European Commission. For the study, 613 websites were analyzed most of which were (454) e-commerce sites and the remaining (159) e-banking sites. The focus of the analysis is on security aspects of payment systems provided on e-commerce websites.
Online payments systems for e-commerce
With the slow uptake of e-commerce, particularly in the busi-
tent on the Internet has grown exponentially over recent
ness-to-consumer (B2C) sector, the initial highly optimistic
years [OECD (2004a)].
growth predictions soon came to be revised downwards. As a consequence of these developments, attention started to
But for these positive trends to finally lead to the long-expect-
focus on the obstacles to e-commerce. Importantly, a strong
ed growth in market share there will be a need for the exis-
barrier to B2C market growth, which surveys continuously
tence of well-developed online payment solutions that are
showed, was strong consumer concern over online payments.
adjusted to the precise payment transaction demands of
At present, with the gain in experience early signs of a trend
e-commerce. A first overview of this financial transfer market
towards increased consumer confidence in online payments
shows a large variety of systems, ranging from credit cards,
can be seen. Another notable development with the gain in
electronic currency, and mobile systems to payment solutions
maturity of e-commerce is the increasing focus of sellers on
via online banking. However, when analyzing actual usage the
providing products suitable to this particular sales channel. In
number of feasible payment platforms, particularly for inter-
particular, the amount of digitally available services and con-
national transactions, decreases to just a few, with a clear lead for credit cards. Thus, with the dominance of credit cards and
100
the success of but a few alternatives, the question arises as to
90
whether the online payment market is prepared to serve
80
e-commerce. In particular, this would demand an ability to
70
allow secure international payments, to conduct person-to-
60
person transfers, as well as payments of different amounts
50
including micro-payments. This article aims to brush some broad conclusions on this question.
40 30
The dominance of credit cards for online payments
20 10
The market has been from the start dominated by traditional
Credit card payment
Direct debit
Fr an ce Lu xe m bu rg Ire Un la ite nd d Ki ng do m
Ita ly Gr ee ce
Sp ai Ne n th er la nd s Po rt ug al Be lg iu m Th e
De nm ar k Fi nl an d Ge rm an y Au st ria
0
E-banking
Figure 1: Payment methods most frequently proposed by e-commerce websites Source: PwC [2003] Factor
Percentage
Concern about security Difficulties to enter information Do not have a credit card Do not like interest charges Purchase value too small Exceeded personal limit
70% 9% 7% 6% 4% 4%
Source: PaymentOne, April 2003
financial intermediaries offering conventional electronic payment services augmented with minor innovations to adapt to the Internet. In 2003, 94.1 percent of all worldwide e-commerce transactions were conducted using credit cards [Pago (2003)]. Furthermore, the current trend suggests that their importance will remain at least as significant in the near future. For example, for the year 2003, Visa Europe registered a doubling of revenues, as compared to the 2002 results [Visa (2004)], from Internet sales to 12.6 billion. A detailed survey of E.U. e-commerce websites2 shows a similar picture, that for these economies classic credit cards are also the most frequently proposed payment systems, with 71 percent. But the survey also shows that other systems are of
Figure 2: Factors discouraging U.S. consumers from using credit cards online
3 For a detailed discussion on security of online payments see PwC (2003). 4 For further details see Visa, 2001a,b and MasterCard, 2003
importance, notably direct debit cards, offered by 45 percent
53
Online payments systems for e-commerce
Additionally, the established trust in banks in some countries
PayPal the seller only needs to indicate the seller’s email
may also allow greater confidence of consumers, as is the
address and the amount to pay. The seller will receive the pay-
case for credit cards [Kallio et al. (2003)]. However, to orga-
ment on his or her PayPal account and will not receive any
nize this system on a national level seems to strongly depend
further financial information about the buyer.
upon banking industry co-ordination and agreements. Given the mainly national organization of the retail banking sector,
PayPal’s success largely came about through its popularity on
the development of international payments is likely to be dif-
Internet auction sites, given that the service permitted per-
ficult and initially costly. Also for micro-payments, it does not
son-to-person payments. Since its purchase by eBay in 2002,
of itself provide an ideal solution.
its aspirations to transform its highly successful U.S. business into a worldwide one have been given a further boost.
Payment means with only limited success — Electronic cash
has an account with the company, a number which corre-
Electronic currency systems, whether in the form of smart
sponds to one-quarter of Citigroup’s clients [The Economist
Currently, one in three online shoppers in the United States
cards or online cash systems, are the electronic equivalent to
(2004)]. Although mostly prominent in online auctions, the
cash. This potentially makes them particularly important from
gained popularity of the service particularly in the United
the point of view of micro-payments and anonymity. However,
States has led to its availability for a range of other payments,
their introduction has largely failed and most of the existing
such as tax transactions. The company is also seeking to
payment means of this kind are not widely used. Their poten-
adjust prices to compete for a share in the growing market for
tial advantages have not been enough to outweigh a number
micro-payments [Navaine (2003)].
of difficulties, such as the question of liability, the cost and way of distribution, and most importantly the difficulty of
Mobile payments: A promise still to be fulfilled
building up a wide network. Furthermore, through ‘virtual’
Mobile payments can refer to a variety of payment systems.
credit card numbers offered free of charge, such as those
For one thing, the mobile device may be used directly for pay-
offered by Citibank, credit cards seem to provide a possible
ment, such as transactions where the telephone bill is
method for consumers to solve the problem. Unless combined
charged. Also, it may be used as a device to pay, for instance,
with other networks, it does not seem that electronic currency
via the bank account. An example of the latter is the Paybox
systems can gain significant market share in the near future
system, which is currently operating in Austria. In this mecha-
as the network constraints seem to be strong.
nism the mobile phone is used as a secure device to permit
Successful alternative payment systems — Mediating services and payments via online banking
the amount to be paid. Confirmation takes place via entering
The mediating service, PayPal, has demonstrated that, under
the customer’s bank account.
payment. The client enters the mobile number together with a personal Paybox PIN. An automatic reply from Paybox acknowledges the payment. The amount will be debited from
favorable conditions, new payment systems can succeed
54 - The
despite the disadvantage of an initial small user base. These
The potential significant advantage of mobile payments is
systems introduce an additional intermediary layer between
that they may provide greater applicability than other pay-
seller and buyer so that payment takes place via the mediat-
ment means because of the very high usage and availability of
ing service provider. To be able to use PayPal, the newcomer
mobile telephones. Mobile devices may thus be developed as
has to create an account with the service, which can be
payment means for bus tickets to offline and online purchas-
charged using, for instance, credit cards. In order to pay via
es. Accepting mobile payments may become attractive for
Journal of financial transformation
5 In a 2002 survey of retailers, conducted by Gartner Consulting, 44 percent of retailers claimed they had new goods and services to sell if a new micro-payment system existed [Maguire (2004)].
Online payments systems for e-commerce
Responding to this problem, credit card companies have
Furthermore, the design of credit card payments is not suitable
developed a series of measures that effectively remove security risks. An important product development to mention is
for transferring small quantities of money. Micro-payments are of growing significance for e-commerce5, particularly because
Verified by Visa. This is a system which connects the card
they are a key condition in the digital content market where a
owner for each transaction directly with the personal bank
range items have low unit costs. The possibility of online rather
through password authentication and a personal message
than physical delivery of items, such as music tracks or digi-
verifying the bank connection. Another important step, which
tized articles, make the electronic sales channel particularly
recognizes that security measures applied by merchants are
attractive. Using credit card payments to realize these small
crucial, has been the move to impose obligatory security standards for merchants4. It remains to be seen whether
payments would render these transactions costly. As a result,
these developments will also lead to increased consumer
confined to subscription type payments. However, as Steve
confidence or whether further initiatives are necessary.
Jobs, CEO of Apple Computer, stated in April 2004: ‘The sub-
Protecting anonymity
music, not rent it’ [Chaffin (2004)].
much of the market for micro-payments does not exist or is
scription services are not succeeding. People want to own their An additional issue related to account-based payments generally is the lack of anonymity. As opposed to transfers involving paper currency, this form of payment allows for the possibility
The contribution of debit cards and online banking
to trace consumer purchases, an issue which raises substan-
In countries such as Germany, Sweden, and Denmark and
tial questions on privacy protection. The current absence of
many other European countries debit cards provide an impor-
online equivalents to cash payments means that consumers
tant payment channel for online payments. One advantage is
leave more ‘traces’ on purchasing habits in the online world
their wider user spread particularly among low-end custom-
than in the offline one.
ers who may not fulfill conditions to obtain credit cards as well as institutional arrangements on credit transactions.
New payment challenges: Micro-payments and auctions
usage of debit cards for international transactions. Because
Agreements with credit card institutions often also allow
It should be noted that credit cards are not suitable for all
the personal account is directly debited some costs savings
payment transfers required for e-commerce. For example, in
over credit cards may arise. However, the basic structure is
online auctions and other markets requiring micro-payments
the same. Consequently, debit cards do not provide solutions
— meaning in particular transfers of small money sums even
to correct the shortcomings of credit cards, such as in the
of less than U.S.$1 — credit cards are not really the most suit-
area of micro-payments. Additionally, debit card payments
able modes of payment. Auction markets, and especially
may benefit from weaker legislative protections [OECD
provider eBay, have gained high popularity on the Internet,
(2002)]. To pay via online banking is a widely used alternative
mainly because they make it possible for private buyers and
payment option in Finland, Portugal, and the Netherlands.
sellers and small merchants to also access large audiences.
Among the various payment systems, the most interesting
However, the characteristics of auctions require the ability to
option is the one where the e-commerce seller offers a pay-
realize person-to-person payments. Sellers wanting to accept
ment option where the buyer is redirected to his or her bank’s
credit card payments need to have a merchant account with
website. The payment is then done using the online banking
credit card companies. The costs involved would not make
facilities provided by the bank. Particular advantage of this
online auction sales worthwhile for a range of small sellers.
system is that existing facilities and security arrangements installed by banks to permit online banking are used.
6 An important step has been taken with the creation of Simpay, a new mobile payment services association, which was founded in February 2003 by Orange, Telefonica Moviles, T-Mobile, and Vodafone. Its aim is to create a network allowing for international mobile payments.
55
Online payments systems for e-commerce
merchants because of the wide spread of mobile phones.
micro-billing solution offers the possibility for consumers to
Wireless access has, as recently stressed by the OECD
pay via their telephone connection.
(2004c), shown the most impressive growth in telecommuni-
There are also prepaid systems providing possibilities for micro-payments on the Internet (e.g. Paysafecard7 and Micro
cation infrastructure. Young people, in particular, which constitute an important part of digital content demand (for
money in Germany). The card is not reloadable and contains
instance, ring tones and games) may not have credit cards, so
no other information than a 16-digit number (PIN), concealed under scratch foil. Other countries, such as New Zealand8,
payment via mobile phone bills (or phone cards) may be the only way for them to purchase goods. Furthermore, it could be
have also developed these kinds of instruments. However, to
devised so as to allow payments from any mobile phone.
the present none of these payment mechanisms have been
Mobile payments may particularly be useful to permit down-
widely offered by providers.
loading and providing mobile content, an area of potentially strongest initial impact.
Another potential solution to the problem of micro-payments has been the development of cumulative collection services.
As for present usage of mobile payments, in the area of digital
Rather than paying for each individual transaction, overall
content related to mobile phones, such as downloadable
expenditure is summed once a month for payment through
games and ring tones, premium SMS is often and frequently
the payment infrastructure. It may be offered by a micro-
used [OECD (2004d)]. However, at present few consumers use
payment organization connecting to a variety of merchants.
their mobile phone as a payment device. For instance, survey
Examples are Cartio Micro-payments and Clickshare. Such
results for Finland find that less than 7 per cent of mobile
payment mechanisms have also been introduced for instance in Germany9 and Denmark10.
phone users used their phone to purchase or order in 2003 (Statistics Finland, Survey Results of Autumn 2003). Also, Paybox, the provider of the most elaborated mobile payment
Growth in the micro-payment options via mobile phones has
system, had to discontinue its operations in Germany, the
been predicted, with ‘direct-to-bill’ expected to be particularly
United Kingdom, and Sweden early 2003, mainly because of
important [Tower Group (2004)]. As for Internet-based solu-
the existing limitations to realize mobile payments.
tions, the aggregation-based solutions are expected to take
Furthermore, developing mobile payments as a means to per-
the lead with prepaid systems being the alternative option.
mit international payments suitable also for micro-payments requires steps to ensure interoperability6.
However, given the current weakness of prepaid solutions (apart from mobile prepaid payments), it is questionable whether prepaid Internet systems will become a powerful
Solutions for micro-payments
means for micro-payments. The future seems to lie rather
Mobile payments have already revealed their usefulness for
with cumulative account systems and mobile solutions. This
micro-payments. An example of micro-payments from tele-
being said, the actual establishment of a leading solution to
phony is the Coinlet system developed by Portalify (Finland),
micro-payments enjoying wide adoption is still to arrive.
which provides for premium-rate SMS and premium-rate voice. The problem with these payments is that they are not
56
Conclusions
widely offered as yet and that frequently are only available on
There are a number of payment systems for online purchase
a national basis. An additional way in which telephony may
that have been proposed and are currently available. At pres-
eventually be successful would be as an alternative way to add
ent credit cards are by far the most popular, with electronic
the cost of the transaction to existing monthly bills (most
currency systems not an efficiently available option. As for
importantly telephone bills.) The German Click & Pay net900
other payment means, PayPal seems to have been able to
7 Paysafecard is a prepaid payment service that can be used for micro-payments on the Internet. The cards are of different values between 25 and 100 and contain a 16-digit number (PIN). 8 Payex is an electronic wallet for online and mobile commerce. 9 Germany has two such systems: FIRSTGATE click & buy and Micromoney.
10 In Denmark, PBS is a service provider for three systems, which operate micro-payments, Valus, EWire, and CoinClick. Internet merchants wanting to accept one of the micro-payment solutions have to make an agreement with one of operators. A customer wanting to make use of micro-payments pays a relatively small amount in advance to the operator using ordinary payment solutions like debit or credit cards.
Emerging models
Six smart moves when playing the smart card game
Leo Van Hove Assistant Professor of Economics, Vrije Universiteit Brussel (Free University of Brussels)
Abstract We are witnessing a shake-out in the European electronic purse market. Several cards that have been in the market for years have or are about to disappear. On the other hand, there are a number of schemes that are doing reasonably well, particularly in the Benelux. This article looks at the mixed experience with e-purses in Europe from a pragmatic point of view and tries to highlight the mistakes that were made by the losers and the smart moves that helped the winners. The article tries to summarize the lessons learned in six 'smart moves'. These lessons should be of interest to newer e-purse schemes elsewhere in the world, but also to other novel payment solutions that target the market for low-value payments.
57
Six smart moves when playing the smart card game
merchants will be reluctant to invest in terminals unless suf-
not very successful on the acquiring side, the number of
ficient consumers have shown an interest, while consumers
GeldKarte terminals remained very low [Van Hove (2004)].
will not use the new means of payment unless there is sufficient merchant acceptance.
With hindsight, a massive roll-out of cards only seems advis-
In order to break this vicious circle, therefore, backers of a
rent promotional efforts on the merchant side can also be
new payment instrument have to succeed, in one way or another, in ensuring that the installed base of both merchants
conducted nation-wide precisely because of the limited geographical scale5. Luxembourg is a case in point here. In addi-
and cardholders exceeds the critical mass point. In the early
tion, chances are that in a small country the number of actors,
stages of the product lifecycle, e-purse operators should,
both issuers of cards and organizations on the acceptance
therefore, focus on building the installed base. Attracting as
side, is lower, thus reducing coordination problems. For exam-
many early adopters as possible, on both sides of the market,
ple, it may be sufficient to convince one or two operators and/
is of crucial importance to get the bandwagon rolling. Indeed,
or authorities in order to attain adequate coverage in, say, the
able in very small countries, where the much-needed concur-
once the adoption rate has surpassed the critical mass point,
parking business. The environment might also be more homo-
the success of an e-purse can become self-reinforcing: the
geneous in terms of the technology used in, say, vending
more consumers use the card, the more merchants will accept
machines.
it, the more interesting its use will become for as yet unconvinced consumers, and so on.
In this respect it is interesting to underline that although Belgium is a relatively small country, the Proton card was
Looking back, most European e-purse sponsors have tried to
nevertheless introduced in phases, city by city. It took Banksys
solve the conundrum by focusing their efforts on consumers
more than two years to cover the whole of the country. The
and hoping that merchants would follow. Obviously, a fast and
French Moneo card — which was launched several year later
massive deployment of cards can be instrumental. It is no
than the first wave of European e-purses, thus providing its
coincidence that some of the more successful European
sponsors with an opportunity to learn from the mistakes of
schemes have succeeded in putting e-purses into the hands
others — was also launched progressively right from the start,
of cardholders very quickly. This is true for miniCASH in
with banks acquiring local merchants as well as installing load-
Luxembourg, Quick in Austria, and Chipknip in the Nether
ing terminals before launching the service with consumers in
lands [Van Hove (2004)]. Rather than issuing stand-alone
a particular region or city.
e-purse cards, issuers in these countries simply incorporated their e-purses into the existing debit cards, and subsequently
Other e-purse operators have turned to geographic focusing
took advantage of a big wave in debit card renewal.
after some time. Europay Switzerland reconsidered its strate-
However, such a ‘big-bang’ strategy is no guarantee for suc-
territory, in 2001 the ‘Zenterstrategie’ was introduced, as is
gy in year 5. Rather than continue to target the whole of the
58 - The
cess. In several countries where issuers flooded the market
evident from the Annual Report for that year: ‘A new concept
with unsolicited cards, many have remained unused. Germany
is geographic focusing: efforts will now be concentrated on
is a salient example. At the end of 2002, some six years after its nationwide launch, the penetration rate of the GeldKarte
so-called ‘A centers’ — or cities, which have converted public transportation to CASH’6. In the same year, the marketing
hovered around 80% of the total population but reportedly
campaign of ZKA in Germany also started targeting cities. In
only 2% to 3% of cardholders were active purse users. Part of
2003, Munich was chosen as the ‘GeldKarte-Vorreiterstadt’
the explanation is clearly that the backers of GeldKarte were
(GeldKarte pioneer city). According to Jan Hendrikx, Euro
Journal of financial transformation
1 Adams, J., 2004, “Leveraging the EMV platform,” European Card Review, 11:1, p. 20 2 Telekurs Group, Annual Report 2003, p. 14 3 Source: European Central Bank, Euro-denominated electronic money in circulation in the euro area, September 2004 4 “25 jaar elektronisch betalen: tijd voor een balans,” Revue bancaire et financière/ Bank- en Financiewezen, Nr. 2004/6, September 2004, p. 302
Six smart moves when playing the smart card game
The second half of the 1990s saw a flood of so-called elec-
and to 240 million in January 2002. The overall amount of
tronic purses — chip cards that can store multipurpose prepaid
e-money in circulation has continued to increase afterwards,
value — being launched on a nationwide scale, especially in
albeit only gradually: the figure for August 2004 is 283 million3. Particularly in the Benelux countries usage of the local
Europe. Currently, however, we are witnessing a shake-out on the European e-purse scene. Several e-purses that have been
e-purses — called Proton, Chipknip, and miniCASH, respec-
in the market for years have either disappeared or are about
tively — jumped significantly following the launch of the com-
to do so. In January 2004, the Swedish banks announced that
mon currency [Van Hove (2004)]. In Belgium, part of the euro
they will close down their Cash scheme by autumn. Four
impact has since evaporated, especially when measured in
months later, Danish EFT operator, PBS, made a similar
terms of active users, and in a recent interview Banksys CEO,
announcement. The Danmønt card, which was launched in
Dirk Syx, called Proton a ‘commercial disappointment’ as transaction volumes remain significantly below target4. Still,
March 1993 and is generally seen as the first real e-purse, will be discontinued from 31 December 2005. Earlier, the
Banksys is not giving up on Proton and will shortly be launch-
Portuguese Multibanco purse — another pioneer — had already
ing an automatic reload function.
de facto disappeared: in 2002 only a few thousand activated cards remained in circulation. Visitors to the Sociedade
So, clearly, the fortunes of European e-purses have been very
Interbancária de Serviços (SIBS) web site, the operator behind
mixed. This article looks at the European experience from a
the Porta Moedas Multibanco, will look in vain for even a minor
pragmatic point of view to find out what can be learned from
reference to a concept that once received tremendous media
these disparate fates and to identify characteristics that can
attention.
help us distinguish between winners and losers. The article
The survival of other purses is also in question. Finland, for
interest to relative newcomers on the e-purse scene outside
puts forward six essential ‘smart moves’, which should be of example, is likely to replace its Avant e-purse with offline use
Europe, such as Dexit in Canada and Edy in Japan, as well as
of a standard EMV debit card, according to Eero Vasenius, Vice
providers of more novel solutions, such as mobile payments,
President of Nordea Bank Finland, on the basis that ‘electronic purse does not fly in Europe’1. The following quote
that to some extent target the markets, such as low-value payments, which e-purses have failed to capture.
taken from the 2003 annual report of the Swiss Telekurs value card CASH has only been managed as a niche product
Smart move number 1: Mind the network externalities
and is no longer being supported in terms of advertising with
Perhaps the most important problem operators are faced with
Group is also not very encouraging: ‘Since the past year, the
large campaigns. The number of transactions conducted with
when launching an electronic purse is the so-called chicken-
CASH decreased in the reporting year by 5.4 percent, to 19.4 million’2.
and-egg deadlock. This is because e-purses are subject to network externalities; put simply, the utility of the payment instrument increases with the size of the network. Stronger
On the other hand, there are a number of European e-purse
still, an e-purse only becomes sufficiently useful if the size of
schemes that are doing reasonably well, thanks in part to the
the network exceeds a certain minimum level, the so-called
introduction of the euro. The aggregated statistics maintained
critical mass. This is true for consumers and merchants alike,
by the ECB show that there was a jump in the value of hard-
although the relevant network differs. What matters for card-
ware-based electronic money in circulation in the euro area
holders is the number of terminals and what is important for
from 170 million in November 2001 to 208 million in
merchants is the number of active cardholders. In such a two-
December 2001, on the eve of the introduction of the euro,
sided market there clearly is a crucial coordination problem:
5 As an aside, this also helps explain why city pilots invariably prove to be more successful than nationwide roll-outs. 6 Telekurs, Annual Report 2001, p. 18.
59
Six smart moves when playing the smart card game
tries where this seems to have happened without problems. In
Smart move number 4: Know your foes
a trade publication, Frans Baeyens, general manager, domestic
Before launching an e-purse it is important to have an idea
payments and deposits at KBC Bank in Belgium was quoted as
about which payment instruments will be its direct competitors
saying: ‘2002 was the first year the banks made a profit on
and how they are priced. This can differ substantially across
this activity. It required an enormous effort, but all the parties
countries. One reason for the lack of success of e-purses in
involved kept the faith in the business case for years on end’12. I have also not come across any evidence of major rifts
Nordic countries appears to be that debit cards can be used for lower-value payments, unlike many other countries. Jyrkönen
between banks in Austria and France, although in the latter
and Paunonen (2003) of the Bank of Finland, for example,
case the national launch of Moneo may have been delayed by
point out that ‘the use of e-money is at a very low level in
the search for a common solution.
Finland. One reason for this is that Finnish consumers are accustomed to paying with other payment cards, especially
In Germany, on the other hand, not all banks have been
w i t h
equally active in promoting the GeldKarte. The cooperative
cards, that can also be used for small payments.’ In this respect
d e b i t
banks, represented by BVR, have added the e-purse to almost
it is interesting to note that, unlike in Belgium for example,
all their ec cards. The Sparkassen-Finanzgruppe, the group of
Finnish debit cards are mainly off-line. Similar explanations
more than 500 savings banks, has also been an active pro-
appear to hold for Sweden, Denmark, and Iceland [Van Hove
moter. The private banks, for their part, have been far less
(2004)]. This obviously reduces the market for e-purses.
active. The Hypovereinsbank, the second largest private bank in Germany (with some 8.5 million clients), has never offered
Where Denmark is concerned, a very specific circumstance is
the GeldKarte. Dresdner Bank decided to stop issuing the
that debit cards are free of charge for consumers and mer-
GeldKarte in June 2002 because of weak demand. Unlike the
chants alike. As a result, according to Henrik Arnt Andersen of
savings banks, Dresdner Bank did not automatically include
Danmarks Nationalbank, ‘debit cards are used for virtually all
the e-purse function on its debit cards. Finally, while the
transactions,’ even transactions of 1 or 2. It is also obvi-
Deutsche Bank 24 (with 7.4 million clients) continues to back
ously important to know how the ‘natural enemy’ of any
the GeldKarte, since the start of 2002 they only offer it as a
e-purse, good old cash, is priced. In many countries, the public
separate card. The result is that, as a survey by InterCard
has been spared the real cost of cash usage. That is, the direct
shows, while on a national scale chip penetration (as a % of
charges faced by consumers when using currency, if any, do
the number of ec cards issued) averages 68%, in some
not cover its full cost. Hence, consumers mistakenly think of
regions this figure is below 26% — depending on which type of
currency as being free. This obviously makes it more difficult
banks have a high market share in those regions. This is particularly true for former Eastern Germany13. With lesson
for e-purses, and other electronic payment systems, to gain ground. In other countries, however, the number of fee charg-
number 1, on the importance of (local) critical mass, in mind,
ing ATMs is on the rise, with the U.K. being one example. All
this is clearly not beneficial for the uptake of e-purses.
other things being equal, in such an environment e-purses should stand a better chance of becoming successful15. Dexit,
In other countries too, some of the issuers have done little more than paying lip service to the common e-purse project14.
for example, appears to be in such a position because Canadian consumers are accustomed to paying for ATM withdrawals.
Summing up, the third major lesson appears to be that the
60
e-purse market is actually a three-sided market, in the sense
Smart move number 5: Know your market
that an e-purse operator not only has to convince consumers
When setting out lesson number 1, it was already pointed out
and merchants, but also needs to mobilize issuers.
that many e-purse operators have piggybacked on the exist-
7 ECR, 2003, “Electronic purse: Last chance for GeldKarte” European Card Review, 10:2, p. 5 8 EURO Kartensysteme, Akzeptanz der GeldKarte in München steigt deutlich Vorreiterstadt-Kampagne erfolgreich, press release, September 4, 2003. 9 “Japan’s contactless cash goes nationwide,” Itworld.com, July 19, 2002; emphasis added.
10 Dexit enables consumers to pay for low-cost items with the tap of a RFID (radio frequency identification) tag linked to a pre-paid account. The RFID tags come in the form of key fobs or adhesive stickers that can be attached to cell phones or other devices. 11 Dexit, Investor Relations, webpage http://investors.dexit.com/overview. asp?ticker=t.dxt, visited on August 12, 2004.
Six smart moves when playing the smart card game
Kartensysteme CEO, such actions are successful; ‘we’ve found that with a push, usage jumps 40%-60% and stays there’7. In Munich, GeldKarte transactions even jumped by 116% in the first half of 20038.
examples he refers to is precisely the battle between Chipknip and Chipper in the Netherlands, where eventually only the Chipknip scheme survived. In April 1999, three years into the ‘standards war’, the two schemes signed an agreement outlining a commitment to ensure interoperability. In March 2001,
Interestingly, e-purse schemes elsewhere in the world already
Postbank and ING Bank decided to abandon their Chipper
seem to have realized the importance of geographic focusing.
technology altogether, phase out the Chipper brand, and
This is how an Internet source describes the strategy followed
migrate to Chipknip by early 2002. This is exactly what the
by Japanese operator BitWallet in promoting its Edy purse: ‘In
theoretical models presented by Leibbrandt predict. The start-
addition to going after large retail chains, BitWallet is also
ing point of his analysis is that payment instruments are sub-
looking to court independent retailers in single locations to
ject to network externalities, as this article has been positing
create something like Edy hot spots where a large number of retailers accept the card’9. There are also indications that
of these, and other, network technologies by existing firms in
all along. In his models, Leibbrandt also analyzed the adoption
Canadian payments firm Dexit Inc., the newest kid on the
an industry, such as banks, as well as the compatibility deci-
e-purse block, is taking that lesson to heart. Established in October 2001, Dexit launched its pay-with-a-wave service10 in
sions that these firms face. He shows that when there is
September 2003 in downtown Toronto. So far Dexit debit
tionally within countries, as is clearly the case with e-purses,
express is only available in the Greater Toronto Area, but
—‘the technology landscape can well be heterogeneous across
Dexit's ‘objective is to expand its service in key strategic areas across Canada, starting in the fall of 2004’11.
countries, [but] it will be homogenous within countries’.
‘autarky’ — meaning that transactions take place dispropor-
Given that, firstly, only one technology is likely to survive and
Smart move number 2: Mind the network externalities
advice is to prevent such a war in the first place. In France and
The second lesson is closely related to the first: in view of the
Switzerland, it initially also looked like several schemes were
secondly, a standards war is harmful for all players, the logical
all-important network externalities, the introduction of several
going to compete head-to-head, but luckily in these countries
incompatible e-purse schemes at more or less the same time
all players agreed upon a common solution before any of the
may well prove detrimental for all of them. This is because
schemes were launched on a national scale [Van Hove
incompatibility creates uncertainty among consumers and
(2004)]. Interestingly, in Finland the first version of the Avant
merchants, and cuts up the market, thus making it even harder
e-purse was launched by a subsidiary of the central bank. By
to attain critical mass. For example, consider the intense com-
taking an active part in the development of the new payment
petition between Chipknip and Chipper in the Netherlands, or
instrument, the Bank of Finland precisely wished ‘to avoid the
the fragmented e-purse market in Spain, where no less than
unnecessary emergence of several parallel or overlapping
three incompatible schemes (Monedero 4B, Visa Cash, and
card systems in Finland’ [Kokkola and Pauli (1994)].
Euro 6000) were launched in a matter of eight months. Tellingly, even today Spain has minimal e-purse usage.
Smart move number 3: Know your friends
In his recent doctoral dissertation, Gottfried Leibbrandt notes
common standard, but for an e-purse to become successful all
Building on the previous lesson, it is one thing to agree on a that ‘there are some interesting examples of firms that have
participating issuers and acquirers must also be completely
tried to compete on standards in payments, but ultimately
committed. This is not evident, as it implies cooperation
joined the industry network’ [Leibbrandt (2004)]. One of the
between competitors. Belgium and Luxembourg are two coun-
12 ECR, 2003, “Electronic purse: Last chance for GeldKarte” European Card Review, 10:2, p. 5 13 InterCard, Chip-Ausstattung der ec-Bankkarten fällt auf unter 70%, press release, April 2, 2003.
14 Examples are the Post Office in Switzerland and Svenska Handelsbanken and S-E Banken in Sweden [Van Hove (2004)]. 15 Although there are currently no e-purse schemes active in the U.K.
61
Six smart moves when playing the smart card game
significant. Compared to cash, however, all the above disad-
pricing?) apply, as do lessons number 5 and 6 in one way or
vantages cancel out. At the same time, e-purses can offer
another. However, there is at least one possible exception:
important improvements over cash. Getting this message
where mobile payments for content or goods purchased via
across may be another reason why e-purse operators should
m-commerce or e-commerce are concerned, there is by defini-
perhaps concentrate, at least initially, on places which previ-
tion less need for a strategy of geographic focusing which has
ously only accepted cash.
proved to be very important in the real world. In short, the caveat is: the ‘smart moves’ put forward here should be trans-
Conclusion
posed in a smart way.
Simplifying somewhat, the ultimate profile of a successful European e-purse scheme seems to be a scheme that is active
References
in a relatively small country and/or one that has opted for a
• Leibbrandt, G., 2004, “Payment instruments and network effects: adoption, harmonization and succession of network technologies across countries,” doctoral dissertation, University of Maastricht, June • Jyrkönen, H. and H. Paunonen, 2003, “Card, Internet and mobile payments in Finland,” Bank of Finland Discussion Paper Number 8/2003, March 12 • Kokkola, T. and R. Pauli, 1994, “Electronic cash,” Bank of Finland Bulletin, 68: 12, December, 9-14 • Van Hove, L., 1999, “Electronic money and the network externalities theory: lessons for real life,” Netnomics, 1:2, 137-171 • Van Hove, L., 2001, “The New York City smart card trial in perspective: a research note,” International Journal of Electronic Commerce, 5:2, Winter, 119-131 • Van Hove, L., 2004, “Electronic purses in Euroland: why do penetration and usage rates differ?” SUERF Studies, Nr. 2004/4 • Van Hove, L., A selected bibliography on electronic purses (and electronic money), web site, 1996-2004
phased introduction, where debit cards are fairly popular but function in on-line mode and/or cannot be used for low-value payments, where all players quickly agreed upon a common solution so that there are no incompatibility problems, where all major banks participate in the scheme and prove to be completely committed, and where the e-purse project receives support from players in at least one and possibly several of the following sectors: public telephones, parking meters, vending machines, and public transport. As already alluded to in the introduction, I am convinced that the recipe for e-purse success elsewhere in the world is similar. However, the analysis of the European experience has also shown that country-specific circumstances can make a world of difference. The single most important lesson is, therefore, perhaps that it is dangerous to generalize. The introduction also claims that the experience with e-purses holds valuable lessons for newer generations of payment schemes. However, while I see clear parallels, there are also important differences. It is, for example, no coincidence that the Dutch banks announced in September of this year that they will stop some of their current individual Internet payment systems — such as Tootz, Way2Pay, and E-wallet — and that they will join forces to develop a new internet payment standard17. The important question is, however, whether a national standard will be sufficient. Turning to mobile payments, it is clear that lessons number 2 (avoid incompatibility), 3 (make sure all GSM operators are involved), 4 (what about
62 - The
Journal of financial transformation
16 A caveat here is that a distinction has to be made between the number of trips registered and the number of actual e-purse payments made (for single fares). Also, Octopus is only slowly making inroads in the retail environment.
Six smart moves when playing the smart card game
ing debit card infrastructure. Incorporating the e-purse func-
operators have been slow in bringing electronic ticketing into
tion into debit cards obviously reduces the issuing costs
their schemes. Part of the explanation for this is that all
involved. It may also be tempting to follow a similar approach
European schemes are contact-based, whereas the preferred
where terminals are concerned, by converting debit card ter-
option for transport applications is obviously a contact-less
minals into so-called 'combi terminals,’ terminals that can
card.
handle both debit cards and e-purses. This has, for example, been done in Austria and in the Netherlands in all probability in an attempt to reach a critical mass of terminals.
Smart move number 6: Know (and educate) your customers In July 2004, the Money & Finance Research Group and the
However, one has to realize that merchants that already
Center for Work, Organizational and Economic Psychology of
accept debit cards are not really part of the core market of an
the Free University of Brussels together supervised a survey
e-purse. As a result, such acceptance points typically do not
of 1,000 Belgian consumers over the age of 15. The survey was
generate high transaction volumes. Rather e-purse operators
commissioned by Bankys, the national EFT operator and spon-
should focus on those retail sectors that typically do not
sor of Proton. The survey confirmed once again that e-purse
accept debit cards, because the average ticket is simply not
users on average tend to be younger and better-educated.
high enough. Also, surveys point out that consumers see the
The same group also proved to be more open to a wide range
added value of e-purses particularly in unattended applica-
of, sometimes hypothetical, payment innovations, including
tions, such as pay phones, vending machines, parking meters,
mobile payments. Early in the lifecycle of a new electronic
and transit systems, because these are the uses for which
payment instrument, when lesson number 1 plays to its full
cash is truly inconvenient.
extent, sponsors might, therefore, want to concentrate on these innovators and early adopters.
An important lesson is, therefore, that besides the quantity of terminals, their quality matters too. More concretely, the
The survey also underpinned the need for thorough, and con-
European experience shows that an electronic purse needs
tinuous, consumer education. Compared to Proton users, non-
the support of at least one and possibly more sectors with a
users consistently underrated Proton acceptance, while over-
large number of small cash payments, such as public tele-
rating the costs involved. Also, more than 8 years after its
phones, parking meters, vending machines and/or public
launch, there were still important misconceptions concerning
transport. Although the importance of public payphones has
the degree of anonymity provided by Proton, with 40.1% of
been diminishing due to the phenomenal uptake of mobile
non-users stating that Proton did not provide adequate levels
phones, the deal with Belgacom early in the lifecycle of Proton
of anonymity, as compared to 23.9% of users. Conversely,
was a vital one for Banksys in Belgium. In the Netherlands,
27.4% of Proton users consider it to be very anonymous com-
purse-only parking in three cities has been, and still is, an
pared to only 15.3% of non-users. In general, as was to be
important driver behind Chipknip usage. Vending machines
expected, non-users were less convinced of the advantages
are typically also popular applications. In the Netherlands, for
offered by Proton. As I have argued earlier, it is important to
example, vending machines represented 5.3% of the terminal
ensure that e-purses are not primarily compared to debit
park at end-2002, but with 18.4% their share in the total num-
cards. Indeed, compared to debit cards, e-purses have obvious
ber of Chipknip transactions over the year 2002 was signifi-
disadvantages. They must be reloaded periodically, there is
cantly larger. As the Octopus scheme in Hong Kong has shown16, another potential volume-booster and door opener
the risk of loss and theft, users must keep track of the balance
might be public transport. However, all in all, European e-purse
advantage of not needing to use a PIN appears to be not so
on their cards, and they lose the float on their money. The
17 “ING staakt Way2pay, ABN stopt E-wallet,� 2004, Planet Internet, September 17
63
Transition
Payments policy in the information age Creating a profitable infrastructure — The payments challenge for banks Payments in transition: Where have all the changes gone? Electronic payments: The drive towards a competitive, customer service driven utility Simulation: A powerful research tool in payment and settlement systems European payment systems and monetary union Technological innovation in retail payments: Key developments and implications for banks
Payments policy in the information age Ronald J. Mann Ben H. & Kitty King Powell Chair in Law, Co-Director, Center for Law, Business & Economics, The University of Texas School of Law
processed and cleared in an electronic way, which justifies rules like those discussed above.
Secondly, to illustrate that framework in application, it is useful to think about the two most rapidly growing payment systems in our economy, credit and debit cards. Existing law
Rules related to error are similar. The types of events that are
draws a stark distinction in legal protections between those
likely to lead to an error, as well as the mechanisms for detect-
two systems. The apparent basis for that distinction is the idea
ing, confirming, and responding to an error are likely to
that cognitive problems that afflict borrowing transactions
depend on the technology that is used to clear and process
justify a greater level of protection for consumers that use
payments. Thus, it makes some sense that the rule for trans-
credit cards than for those that use debit cards. From that
actions processed electronically (covered by the EFTA) would
perspective, the existing rules divide consumer payments into
differ from the rule for transactions processed entirely by
two classes. The first class includes transactions in which the
paper (conventional check transactions governed by Article
consumer makes contemporaneous payment: cash, checks,
4). At the same time, the continuing shift of check transac-
and debit cards. Because the consumer at the time of the
tions from paper to electronic processing (probably to be
transaction understands that the payment is being made
accelerated by the Check 21 Act) might undermine that dis-
more or less immediately, the consumer is treated as ade-
tinction.
quately assessing the wisdom of the payment in question.
Rules that determine when a payment is made are similar, in
Credit cards, however, are quite different from that perspec-
that they are for the most part made based on the practicali-
tive, because the consumer that purchases with a credit card
ties of a particular system. Thus, in the wire-transfer system,
does not make immediate payment. Rather, although the mer-
we say that the payment is complete when the beneficiary’s
chant receives contemporaneous payment, the payment by
bank becomes obligated to pay the beneficiary. In the check-
the consumer is deferred, automatically until a statement is
ing system, we say that the payment is not complete with
received and, at the consumer’s option, more or less indefi-
respect to an ordinary check until the check is paid, but that
nitely, as permitted by the strikingly lenient repayment
it occurs with respect to a cashier’s check when the payee
options typical of the modern American credit card.
accepts the instrument. The Anglo-American legal system has a long tradition of proRules related to reversibility, however, are completely differ-
tecting borrowers from the folly of imprudent borrowing. The
ent. Rules related to reversibility should depend on the
most famous example surely is the centuries-long effort of
dynamics of the underlying transaction in which the payment
English courts to invalidate a series of creditor devices that
is made. In the simplest cases, payment systems are special-
had the effect of granting mortgage creditors a broad right to
ized for use in particular situations. Thus, for example, in busi-
take real-estate collateral from borrowers who failed to per-
ness transactions, parties often choose to make payments
form precisely as they had promised at the time of the loan.
with letters of credit or wire transfers. Those systems include
That instinct continues to have broad application today, as
particular rules designed for the particular transactions in
courts steadily broaden the range of devices to which that
which they are used, which determine the timing and circum-
invalidating rule extends.
stances in which payments can be recovered or stopped once
66 - The
the process has been initiated. Because those systems are
Although it is a bit much to superimpose the insights of mod-
quite specialized, the system-specific rules work well for them.
ern academic literature on the policy instincts of the medieval
Journal of financial transformation
The point of the longer essay this paper summarizes is to
At its heart, payments law must resolve four fundamental
emphasize something that is deeply wrong with the current
questions: who bears the risk of unauthorized payments, what
framework of our payments policy. The portion of that policy
must be done about claims of error, when are payments com-
reflected in the Truth in Lending Act grants consumers a vari-
pleted (so that they discharge the underlying liability), and
ety of generous rights in credit card transactions. By contrast,
when can they be reversed. The first three questions are cat-
the Electronic Funds Transfer Act, which governs debit card
egorically different from the last, because they often should
transactions, is relatively chary in its protections. Today, how-
be resolved based on the nature of the underlying technology.
ever, with the debit card market increasingly dominated by the
Thus, for example, with respect to the risk of unauthorized
PIN-less debit cards marketed by Visa and MasterCard, the
payments, the fundamental question is how to design a sys-
distinction between the credit card and the debit card is
tem that gives adequate incentive to the user to avoid and
almost invisible to all but the most sophisticated consumers.
mitigate losses from unauthorized transactions, while giving
The cause of the problem is easy to see, of course: technology
in technology and system design that can avoid and mitigate
adequate incentive to the system operator to make advances has altered the landscape of private payment institutions
those losses. In our legal system, we have taken the view for
while Congress has not had occasion to update the statutes
most high-technology payments that an almost complete allo-
that regulate those transactions. The solution, however,
cation of the risk of those losses to the system operator is
seems sufficiently difficult to warrant detailed consideration. I
appropriate.
start at a high level of generality, analyzing the general question of what types of considerations should inform a sophisti-
The premise of those rules (admittedly unspoken) is that even
cated payments policy. There is of course some substantial
a complete allocation of loss to the network operator will
prior thinking on the subject, most obviously the Uniform New
leave the consumer a sufficient incentive to attend to these
Payments Code (UNPC) and writers like Peter Alces who criti-
problems. That could be true because of the hassle of revers-
cized that project. I want to make three major points here.
ing unauthorized charges, because of doubts that financial
The first point is that previous analysis has failed to recognize
tion, or even because of ignorance of the legal protections for
institutions will readily fulfill their obligations in such a situathe importance of the underlying transactions in which pay-
unauthorized transactions. At the same time, the rules reflect
ments are made to issues ordinarily treated in the legal rules
the implicit premise that losses in technology-driven systems
that regulate payment systems. The basic problem is that pay-
are most effectively reduced by technological and system-
ments policy needs to attend more consciously to the con-
design initiatives that are exclusively within the control of the
texts of the transactions in which payments are made.
system operator. Thus, we are not surprised to see major
Existing law articulates rules that are bounded almost entirely
investments in fraud-prevention technology in the credit-card
by the nature of the technology with which the payment is
and debit-card sectors. Because the justifications for those
made. Thus, we have separate rules for wire transfers, letters
rules relate to the nature of the technology, it is plausible for
of credit, checks, electronic transfers, and the like. That type
federal law to prescribe such a rule for all electronic transfers
of boundary makes sense only for issues driven by the nature
from consumer accounts. It is less plausible to include a simi-
of the technology. It makes no sense, however, for issues that
lar rule for credit card transactions based on the availability of
should be resolved by reference to the nature of the underly-
credit in the transaction. It would be more sensible, surely, for
ing transaction in which the payment is made.
that rule to be justified by the fact that the transactions are
67
monthly statement, the task of monitoring expenses to review
not. In the context of face-to-face transactions, I conclude that
compliance with a budget (or, perhaps more commonly, to
credit and debit cards should have similar limitations on
understand the failure to comply with a budget) is rendered
reversibility, leaving open the possibility that developing for-
much easier.
mats like prepaid cards ultimately might be left on the cashequivalent side of the line. In the Internet context, where
Still another factor, at least by comparison to the check, is the
credit cards currently dominate, I conclude that the same rule
desire to close the transaction rapidly. Although this has not
should apply. To be sure, that does limit the availability under
always been true, it now plainly is the case that the check is
current technology of any payment option in that milieu that
the slowest of retail payment mechanisms. When a grocery-
is both practical and wholly final. In the end, however, the
store customer pulls out a check-book to pay for groceries,
distinctions between credit and debit cards do not seem to me
the hurried customers in the line behind inevitably sigh, know-
sufficient, or sufficiently clear to the user, to justify a major
ing that their wait will be protracted by the additional time for
distinction on that point.
the consumer to write the check and for the clerk to decide whether to accept it. That is not to say, of course, that the use of credit cards for borrowing is not an important policy issue. It is to say, however, that it is not sufficiently dominant in the decision to use a credit card to exclude other relevant concerns. That argument leads to, at least, two possible responses. The simplest course would be to repeal the TILA protections as outdated. I reject that approach. Instead, I build on the framework articulated above to argue that those rules have come fortuitously to serve other important policies, specifically the transaction-related sales policies that are important in issues of reversibility. The most obvious one is to redress the imbalance in enforcement capabilities between the typical merchant seller and the typical consumer buyer. If the existing rules shift the burden of instituting and pursuing litigation from the consumer to the merchant, they might enhance the likelihood that disputes in retail transactions are resolved appropriately. If that is true, then it justifies the TILA rules. The problem with that justification, at least as an explanation of existing law, is that the enforcement-imbalance concern applies to any payment system that the consumer uses. The major point cutting against broad application of that rule is the need to maintain payment options with different characteristics, so that merchants and their customers can choose between payment systems that are final and those that are
68 - The
Journal of financial transformation
English judiciary, that policy finds broad support in the spe-
the credit aspect of the purchase transaction is increasingly
cific concerns of the nascent behavioral economics move-
obscure. For many of us, probably most of the limited sample
ment. Scholars in that field often point to an overlapping set
of people who will read this article, the credit card is used
of tendencies that generally lead a normal individual to under-
much more for convenience than for any purpose related to
estimate the likelihood of negative future events that have not
borrowing. To put it another way, we often pull a credit card
previously been salient in the experience of the individual’s
from our wallet not because we lack the present wealth to pay
circle of personal acquaintances. A similar, related phenome-
the purchase price for the item in question, but for some other
non leads to systematic underestimation of the likelihood that
reason unrelated to a credit decision.
a negative event will happen to the estimating individual, even if the individual accurately understands the overall likelihood
If there is anything that proves that point, it is the rise of the
of the event. Both of those phenomena are exacerbated by
convenience credit card user, who charges amounts on the
the likelihood that consumers have higher discount rates for
card but pays off the bill each month. Once a relatively small
events perceived as likely to occur far in the future than they
sector of the market, convenience users now constitute about
do for events likely to occur in the immediate future. Thus,
40% of the market. For those users who consciously limit
those scholars would say, it is reasonable to worry that bor-
their spending to what they can repay out of resources avail-
rowers entering into credit transactions do not adequately
able that month it is almost deceitful to treat the transaction
weigh the likely harms to them from the difficulties that might
as one involving a significant extension of credit.
come at the time that repayment is due. Also, with a little more subtlety, those who pay in advance might underestimate
Rather, we use the credit card for a locus of completely differ-
the likelihood that they will harm their strategic relations with
ent reasons. The most economically rational might point to a
the merchant by agreeing to pay now and receive the subject
desire to get the float, the use of the funds during the time
merchandise (or services) later.
that elapses between the date of the transaction and the date that the credit card bill must be paid to avoid a finance charge.
Finally, that concern works well as a lens for understanding
But the rapid rise of the debit card, which has no such float,
TILA. TILA does little or nothing to regulate anything about
suggests that it is not the dominating market factor in the
credit card transactions divorced from their credit-related
choice of a credit card over cash or a check.
features. The bulk of its regulations appear to rest on the concern that crafty credit card issuers will trick consumers
More likely factors include such things as a desire not to have
into using credit cards without understanding the costs of
to carry enough cash to cover all of the purchase transactions
repayment. Thus, the statute relies heavily on a variety of
of our daily life. That is not just a desire to limit the risk of a
disclosure rules reflecting the well-intentioned notion that
loss from theft. It also has to do with the relative ease with
mandatory disclosures will resolve the cognitive problems
which credit cards can be used, by comparison to the ease
that afflict the potential credit card user.
with which we can get cash from our banks: how many of us stand at retail counters thinking something like ‘if I use a
Thirdly, it seems clear that the normative underpinnings of
credit card here it means I won’t have to stop at the ATM to
the current rules have become outmoded in several ways.
get money before I go to work tomorrow?’
Most importantly, it seems likely that many, if not most, credit card transactions no longer reflect the kinds of long-term bor-
Another consideration for some of us relates to the ease of
rowing transactions that justify the cognitive concerns that
record keeping. If most of our transactions can be pushed
would justify the existing policy. In the modern era, however,
onto a single credit card, which provides a single consolidated
69
Creating a profitable infrastructure The payments challenge for banks Ann Cairns Global Head, Working Capital, ABN AMRO
route favored by some scale players, such as Citibank. But
lenge facing any other business. In every business there are
banks will need to make their own assessments of which best
three competing business themes, customer-relationship cen-
fits their own institution.
tric, product-innovation centric, or infrastructure manage-
Offshoring is here to stay
publishing, etc) offer examples of how organizations have
ment centric. Different industry sectors (pharmaceuticals, Part and parcel of organizational decisions are choices about
evolved to unbundle conflicting core processes. The banking
whether or not to offshore. Whether institutions have the
and the payments business is no different. It too will consist of
capability to execute offshoring within their existing network or not, an offshoring2 strategy is a critical component of any
flict or be suboptimal. The challenge is to choose which of the
these three unique business characteristics which may con-
payments strategy where quality, productivity improvement,
three models will add maximum value to clients and share-
and reduced manufacturing cost are key competitive factors.
holders in the context of the strategy of the parent institution.
The direct benefits of offshoring are clear: Up to 65% reduc-
A customer-relationship business
tion in payment instrument unit processing cost, with signifi-
Role: identify, attract, and build relationships with customers.
cant quality enhancements generating downstream benefits
Tasks: draw customers into the bank, assist them, and try to
in customer satisfaction and reduced service and corrective
build personal relationships with them, responding to ques-
costs. Additional benefits come from the enormous catalytic
tions and complaints, collecting customer information, etc.
effect of re-engineering processes that enable a relentless pursuit of continuous quality improvement.
A product-innovation business Role: conceive of attractive new products and services and
Some banks, including ABN AMRO, have been able to signifi-
commercialize them.
cantly enhance disaster recovery and business continuity by
Tasks: research new products, figure out how to bring new
introducing ‘load-sharing’ between regional and offshore
products/services best to the market, find ways to present
processing hubs. The reduced risk of disruption is illustrated
new products to potential customers, etc.
by the fact that four power outage and security related site denials in Western locations resulted in zero customer
An infrastructure business
impact, due to the ability of the offshore facility to execute
Role: build and manage facilities for high-volume, repetitive
processing. Counterintuitive to a ‘West is best’ outlook, per-
operational tasks.
haps, but true nevertheless. In fact, the key offshoring deci-
Tasks: build new branches, maintain data networks, provide
sion for banks is not whether to offshore, but whether to do
back-office transactional services, manage logistic networks,
so in-house, by building their own facilities, or to select a
etc.3
trusted provider — either a local company or a Regulated Financial Institution with a proven business process outsourc-
Excelling at all three at the same time is incredibly difficult
ing offer.
and results in suboptimal contribution to the parent institution. For example, when we explore the economics we find
A payments model appropriate to the parent institution’s strategy
that economies of scope apply to the customer relationship
Payments business groups within banks need to develop strat-
ship maintenance makes it an imperative to gain large share
egies that maximize value for their parent. In essence, the
of wallet, economies of speed apply to product innovation
payments challenge is no different from the commercial chal-
aspects of the business where rapid time to market equates to
theme where high cost of customer acquisition and relation-
1 Article in Computer Weekly, 13 July 2004
70 - The
Journal of financial transformation
On the face of it, the payments industry is thriving as volumes
ments organizations, according to market estimates. The
continue to rise steeply. In 2003 NACHA processed 10 billion
global IT investments (i.e. not just payments) of financial
payments (up 12% on 2002), SWIFT volumes grew by 9.5% on
institutions are an estimated U.S.$347 billion, according to
2002 (payments represent 60% of all SWIFT messages), and
Tower Group (2004). At the same time, the costs of meeting
The Bank for International Settlements reported daily global
regulatory requirements have spiraled. Know your customer/
turnover in traditional foreign exchange markets of U.S.$1.9
U.S.A. Patriot Act, Basel II, and Sarbanes-Oxley together rep-
trillion in April 2004, up 36% at constant exchange rates com-
resent significant costs. HSBC, for example, has calculated
pared to 2001.
that the cost of meeting regulatory requirements is 3.125% of
But, as every payments executive knows, despite rising vol-
amortization)1.
its Ebitda (earnings before interest, tax, depreciation, and umes overall, the industry actually faces something of a crisis. Downward pressure on pricing and the costs of doing business
The high costs of payments infrastructure have been com-
in an increasingly competitive and regulated market place add
pounded in most markets by low interest rates (on float) over
up to a very challenging environment for payments providers
recent years, adding to the difficulties for individual payment
in 2005.
providers of supporting their payments infrastructures. While
Downward pressure on pricing
others are certainly with us for the longer term.
this cost factor looks likely to improve going forward, the Payments are, fundamentally, a utility users are reluctant to pay for. So-called ‘free banking’ has been the lifetime experi-
Scale and size are different
ence of most adults in the retail banking sector in, for exam-
The received wisdom over recent years has been that market
ple, the Netherlands and the U.K. Businesses too have become
consolidation will resolve these challenges through the emer-
much more demanding buyers of payment services over
gence of a few mega providers able to process high volumes
recent years; reasonably enough, they negotiate hard with
of payments efficiently and profitably. Yet mergers, especially
their bankers on pricing and they also manage their payments
cross-border, are hard to achieve and difficult in execution. It
needs closely using such tools as inter-company netting and
may be several years before benefits start to be realized. And
domestic non-urgent payment methods in preference to
even if the desired scale can be acquired in this way, will the
cross-border transfers, and so on.
organizational structure be able to exploit it efficiently? As anyone who has spent more than five minutes in a very large
And gradual downward pressure on pricing has been sharply
organization knows, big is not always beautiful and the day-
accelerated in Europe by the introduction of the Single
to-day realities of siloed businesses, fragmented infrastruc-
European Payments Area. Within the Euro-zone, cross-border
ture, and allocated overheads can quickly dissipate the ‘scale
transfers of up to 12,500 must now be charged as domestic
benefits’ of the text books.
transfers — the cap will increase to 50,000 in 2006. So getting the organizational and sales models right will be
Cost drivers — technology renewal, market complexities, and regulation
ume business. Boston Consulting Group’s research has done
Meanwhile, the costs of providing the reliable and compre-
much to inform thinking in this area. Of three organizational
every bit as important to profitability as assembling the vol-
hensive payments infrastructure demanded by users contin-
models identified by BCG (infrastructure, customer centric,
ue to rise. Technology costs represent a huge and ongoing
and transaction) the transaction model (a mono-line, custom-
burden, averaging around 20% of the total cost base of pay-
er-facing entity with full P&L responsibility) appears to be the
2 For more information about offshoring please refer to the 8th issue of this Journal, which was dedicated to this topic. 3 This concept, which is evident in our business, is identified and developed by Hagel, J. and M. Singer, 1999, Net Worth: Shaping Markets When Customers Make the Rules, HBS Press
71
premium pricing and larger market share, and economies of
Bank moved all domestic payments to Postbank in 2003, giv-
scale apply to any infrastructure business where high fixed
ing Postbank a 15% share of the German domestic payments
costs require large volume throughput to drive low unit cost.
market. Then earlier this year, Deutsche Bank took over handling of all Postbank’s cross-border business.
In or out: What is a bank to do? While choice of organizational model will be critical to existing
In the future, we can expect to see banks and financial institu-
or aspiring core infrastructure providers, other banks may
tions pursuing still more innovative organizational models as
make a different fundamental choice — that of outsourcing or
the market segments into core infrastructure providers and
partnering to resource the payments services the bank and its
those financial institutions that choose to outsource their pay-
customers need.
ments processing. For both, the challenge will be to maintain organizational clarity and flexibility, and customer responsive-
In making this decision, payments executives will obviously be
ness, as they strive to become or remain profitable.
looking to shed the infrastructure costs referred to above, but they should also be thinking positively about the best way to enhance their core capabilities and extend their reach, as well as manage growth and risk for the future. Whichever path an institution takes, the decision point is now (or earlier!). Some banks are already incurring significant losses across their payments franchises, decisions made now concerning infrastructure changes will take some years for the benefits to be realized, and banks that hesitate will be even more uncompetitive once earlier adopters of new business models start to reap the benefits of, for example, new technology platforms or offshoring models. Those banks choosing not to be infrastructure providers still have complex decisions to make. There is a spectrum of choice from outsourcing a specific service for a specific market segment or geography (for example, a bank may choose to outsource all euro clearing but continue to clear dollars themselves) to fully private-labeled outsourced structures (for example, front- and back-end processing for all corporate payments). In the latter case, the bank would concentrate on providing an effective sales coverage approach for payments as an add-on to other, value-added financial services. In other cases, strategic partnerships may be founded on mutual benefits, as with the partnership between Deutsche Bank and Postbank4. In this interesting alliance, Deutsche
4 Discussed in more detail by Wolfgang Gaertner in this issue of the Journal.
72 - The
Journal of financial transformation
Payments in transition: Where have all the changes gone? Mark Webster Partner, Capco For many years, pundits have been predicting the dawn of a
cussed later, this situation is exacerbated by the fact that few
cashless and checkless world and the demise of more tradi-
banks look at payments holistically. Most banks operate in
tional transaction methods. In the U.S. specifically, the ‘death
silos and have created separate operational entities for paper
of the check’ has been heralded and discussed for at least the
and electronic payments. This makes it difficult to determine
last thirty years. The reality is, however, that the uptake of
the overall economic impact of changes in transaction pat-
new and improved payments products has been much slower
terns and results in continued investment in existing systems
than anticipated. Paper-based manual payments systems are
that have already reached scale economy, rather than invest-
still prevalent and most banks are still using payments plat-
ing in newer unproven payments vehicles and platforms. This
forms and architectures that had their genesis in the 1960s
situation is changing as maintenance costs for aging pay-
and 70s. Given the broad spread acceptance of personal com-
ments platforms increases. But at the moment, banks tend to
puters and Internet communication, the question becomes
be more focused on increasing the cost efficiency of their
why has the payments industry been so slow to change?
existing payments products rather than making extensive investments in new systems.
Drivers of change By their nature, financial institutions are inherently conserva-
The economic situation is complicated by the surprising lack
tive and risk averse. Barring regulatory demand, change usu-
of customer demand for newer payments products. Con
ally takes place very slowly, almost coming to a complete halt
sumers have been very slow to move from paper to electronic
without some strong motivation. The typical drivers for
payments. Despite a variety of marketing campaigns by banks,
change in banking tend to be cost control or efficiency, cus-
retailers, and trade associations, it has only been in the last
tomer demand (revenue pressure), and risk control. While
year that the volume of retail electronic payments, primarily
each of these areas can be affected by transformation of pay-
card and ACH, has equaled that of the paper-based market in
ments systems, the overall impact of these drivers has been
the U.S. Businesses have been even slower to change their
less than impressive, in terms of mandating change.
payment habits. Trade organizations, such as the Association of Financial Professionals (formerly the Treasury Management
The foundations of much of the current payments system
Association), have made it very clear that checks are the pre-
infrastructure have been in place since the 1960s. As business
ferred method of payment, at least from the disbursement
and consumer demand for financial transactions increased in
side of the house. This lack of demand has made it difficult for
a post-World War II era the public and private sectors, in the
banks to justify huge investments in new payments infrastruc-
form of banks and central banks, joined together to create an
ture. Faced with taking the risk that ‘if you build it, they will
incredibly large and sophisticated infrastructure to manage
come,’ banks have again focused on improving the overall
the flow of paper-based transactions, primarily checks. As
efficiency of traditional payments platforms, and only making
volumes grew, existing processes were automated to improve
limited investments in newer systems.
efficiency and entire businesses grew up around meeting the needs of the check-processing industry. The end result has
The third potential driver of change is risk control. This can
been the creation of a large base of fixed costs dedicated to
be looked at from several points of view. Firstly, it is the issue
processing manual transactions. Electronic transactions are
of transactional risk control. Banks and corporations have
arguably more cost effective than corresponding paper-based
spent significant time and money developing technology and
ones, on a per transaction basis. However, when the costs of
processes to control and manage the risk of check fraud.
supporting the existing infrastructure are included, the busi-
Although check fraud is a growing concern for most banks
ness case for change becomes questionable. As will be dis-
and many corporations, few consumers see check fraud as a
4 Discussed in more detail by Wolfgang Gaertner in this issue of the Journal.
73
major issue. By comparison, both banks and consumers are
resist change unless there is some strong incentive to make
increasingly concerned about credit and debit card fraud.
the change. In the case of payments, one might expect this to
Banks have spent large amounts of time and money develop-
be an economic incentive, given that electronic payments
ing ways to control the risk of card fraud, but losses continue
should be cheaper and more convenient than the correspond-
to climb and consumers’ fears, largely due to growing con-
ing check-based ones. While this might be a true statement on
cern about identity theft, continue to increase. As a result, at
an all-in cost basis, most customers, whether consumers or
least for the moment, transactional risk does not seem to be
corporations, do not see it this way, mainly because there is
an adequate reason to radically change payments systems.
lack of transparency in payments processing costs in the
The second point of view is that of operational risk. The
banking industry. Very few bank customers see the actual cost
increasing cost of maintaining existing payments platforms
of check writing as a disbursement method. Rather than
may be an incentive to move towards newer and more effi-
charging transaction fees, most banks recover their costs
cient platforms. As previously mentioned, much of the exist-
through a combination of float and exception item service
ing payments infrastructure has been in place for thirty to
charges for check writers and deposit charges levied on com-
forty years. Although the systems and technology have been
mercial customers. In fact, many U.S. banks have extensive ad
updated and modified over time, their overall complexity and
campaigns around variations on the theme of totally free
the related cost of maintenance continues to grow. Despite
checking. Even for commercial customers, the per-transaction
the increasing cost, however, the existing systems do work,
or ‘penny price’ for individual checks is rarely equal to its true
and as the old adage says ‘if it ain’t broke, don’t fix it.’
cost. This apparent subsidy of check writing by the banking industry provides customers a strong incentive to stay with
Barriers to change
their traditional habits of writing checks, while check receivers
While the drivers for change may be mixed, barriers prevent-
bury their costs in the prices that they charge the end-users.
ing it are significant. Here again we can look at risk, revenue or customer demand, and cost efficiency. In each case, there
This lack of transparency is also an issue within most banks.
are factors that create potential barriers to any change bar-
As previously mentioned, most banks are highly siloed when it
ring absolute necessity.
comes to payments platforms. Very few banks have the ability to look at payments as a whole and have difficulty apportion-
Payments are a core function for financial institutions.
ing costs and revenues across various payment channels and
Various studies have indicated that payments account for
platforms. As a result, decisions are often made based on an
30% to 40% of total net banking revenue. Understandably,
incomplete understanding of the impact of change on all pay-
most banks are hesitant to make major changes to core sys-
ments products. While free checking may be viewed as a
tems in a big bang approach when so much of their revenue is
needed prerequisite to maintaining demand account balances,
at stake. The risk of potential failure is just too high. Unfortun
electronic transactions are fully charged, resulting in an
ately, the same complexity and age of existing systems that
added customer disincentive to change. Additionally, invest-
increases maintenance costs also makes it difficult to make
ment decisions are often made based upon the relative size
significant changes incrementally. As a result, major changes
and importance of the payment channel, rather than on the
in existing systems tend to be deferred if at all possible.
long-term strategic interests of the financial institution. This tends to favor existing payments systems and platforms that
Although referred to earlier as surprising, the lack of cus-
have already developed scale volume.
tomer demand for new and improved payments systems is actually not that surprising at all. Most consumers tend to
74 - The
Journal of financial transformation
Finally, from a cost efficiency point of view, payments process-
ing is a highly complex network system. This is especially true in the U.S., since there are a large number of financial institutions that clear payments. In large systems of this type, participants are highly dependent upon the actions of other network participants. One bank cannot change to a more efficient method of processing payments unless a significant number of the other participants make a similar change. As a result, in a variation on the traditional concept of ‘The Prisoner’s Dilemma’ in gaming theory, the network tends to suboptimize and discourage change. Since everyone can process checks, we continue to process them and there is little movement towards more efficient payment methods.
Conclusion While the economic transparency of our various payments systems is unlikely to improve any time soon, end-users are finding other reasons to demand incremental, if not overnight, change. Decades of such incremental changes have already had significant effects on both check and electronic payments systems. Over time, we have gradually built an electronic infrastructure that can effectively parallel existing paperbased systems. The increasing age of the existing systems will eventually force a transition as maintenance and operating costs continue to increase. Regulatory issues, such as Check 21 in the U.S., image processing mandates in parts of AsiaPac, and pricing harmonization in the E.U., will only accelerate the rate of change. We may never see a totally cashless and checkless society, but the transition to more integrated and automated payments systems is happening. Rather than the revolution in payments that was once predicted, we are in the midst of an evolution, a slow sea change that will occur without most of us ever realizing that there has even been a change. How we get there is perhaps not as important as the fact that we will eventually get there. And by then, the pundits will be predicting even more radical changes!
75
Electronic payments: The drive towards a competitive, customer service driven utility Rod Dew Marketing Director, Distra Pty Ltd
Once upon a time a Bank CEO received a phone call from his
dramatically in that period. Why not? The bottom line is that
mother on Christmas Eve.
payments have to be fulfilled as a core mechanism in our economies. The risks, scale, security, and complexity cannot
— ‘I can’t get any money out from the ATM for my last min-
be underestimated in delivering a successful payments net-
ute shopping, the Bank branch has long queues and
work with appropriate regulation to ensure that the necessary
I couldn’t use my Bank card at the supermarket,’ she
robustness, safety mechanisms, and competitive frameworks
exclaimed.
are in place. The inertia to change and the high barriers to
— ‘Don’t worry, I’ll look into it’, he replied, thinking this was a
entry have been dictated by sophisticated groupings of inter-
minor problem as he prepared for his annual talk to the
ests, including governments and regulatory bodies, financial
senior executives over sherry and mince pies.
institutions, large payments processors, large retailers, consumer groups, and service and technology suppliers.
He called the operations center; ‘Actually the Bank’s whole ATM and POS network is out of operation and it is unlikely that
The above story actually illustrates only one aspect of the
we can recover the situation for another two hours,’ reported
increasing pressures on payments processing to deliver
the duty manager, lump in throat. The implications almost do
improved and new services at a significantly lower cost. It is
not bear thinking about; the CEO knew that 25% of the entire
possible that this Bank had driven costs down through low
country’s electronic payments network was out of action on
investment in aging infrastructure, asset depreciation, and
the busiest shopping day of the year.
staff reductions. As transaction volumes have increased and layer upon layer of complexity introduced, the system has
Why did the CEO not know? What are the ramifications from
reached its capacity and failed.
the merchant business customers and card holding clients? How long and how much will it cost to handle the manual
Payments processors are facing a difficult equation. They
transactions generated from the outage? What does this
have to assess whether the commercial risks of not changing
mean to the Bank’s reputation and brand? This is a business
their business processes and infrastructure are greater than
whose raison d’être has been based on technology to deliver
the risks associated with technological change in one of the
convenience; even his mother relies on electronic payments
most demanding of all IT and telecommunications environ-
mechanisms.
ments.
How many times have we heard it? ‘This business needs to
Outsourcing remains another option where banks have con-
improve client service levels, adapt to changing circumstanc-
tracted either specialist third-party payments processors or
es, increase revenue, and reduce costs.’ The financial services
major outsourcing organizations to operate their IT infra-
industry is littered with examples of not heeding to or execut-
structure. Evidence suggests that these outsource operations
ing successfully around these core facets. For example, new
actually have the same problems as the original payments
entrants in mortgages, on-line share trading, and insurance in
institution, namely; limited differentiation, rigid processes,
the last ten years have changed the market permanently. The
ageing infrastructure, and costs levels that do not support
top ten in many lists of vertical product suppliers will have
new market models and competitive dynamics.
changed beyond recognition in the last decade in most developed countries.
Whilst it is relatively straightforward to argue that a bank’s core business may not be operating a Point of Sale (POS)
However, payments have not been tested and threatened as
76 - The
Journal of financial transformation
terminal fleet, the advantages to the bank include: Revenue/
profitability, a direct touch point with the business customer,
ments network, there are a number of other forces of change
service differentiation, and cross-selling opportunities. It is
that are leading to a re-assessment of the payments business.
possible that outsourced or central switch initiatives could
The following provides a summary:
dilute the bank’s brand and disintermediate the bank’s relationship with the merchant customer, as witnessed in the
Regulation — an increasing number of regulations are impact-
United States and by smaller financial institutions that utilize
ing payments processing globally. These include Check 21
outsourced payments services.
(check image capture at the Point of Sale in the U.S.), the Patriot Act (determining disclosure obligations on money
Merchants demanding more
laundering), Basle II (operation risk capital adequacy — as per
Merchants have become increasingly dissatisfied with the
the story referred in this paper), reduced payments network
level of service offered by the payments network. Currently, a
fees, and greater transparency (Reserve Bank of Australia
typical model includes paying rental for the Point of Sale
credit card reforms and EFTPOS network designation, and the
devices and telecommunications line, and a fee for each trans-
Treasury Select Committee enquiries in the U.K.).
action processed. The transaction fees cover both the processing and support costs of the acquiring operation and fraud and other risks faced by the card issuer.
Role of card schemes — potential central switching models in Asian countries aim to circumvent the perceived expense of utilizing the card schemes as switch processors. Additionally,
Settlement, the movement of funds from the cardholder’s
U.S. merchants have fought successfully for the lowering of
account to that of the merchant, can take days, and transac-
debit card transaction fees for transactions processed through
tions that are disputed can take weeks to resolve. The e-pay-
the card scheme networks.
ments POS devices are generally not integrated with in-house networks, inventory and financial management systems.
Fraud — increasing card fraud has led to national mandates to deploy smart card technology (EMV). This typically requires
These factors have led to high rates of churn at the small
the deployment of new terminals and card readers, and
retailer end of the market, driven by marginal fee benefits,
changes to the core processing software.
and for larger retailers to deploy their own infrastructure (or retailer led outsourcing/ consortiums) to negotiate better
New channels and transactions — growth in ATM and POS
transaction fee terms and lower infrastructure costs.
usage continues unabated. However, newer channels, such as
An opportunity exists for enterprising payments processors
els of market acceptance. POS devices can also be extended
and merchant acquiring banks to grow market share by provid-
to provide banking services, such as funds transfer and bill
the internet and mobile networks, are gaining significant lev-
ing greater information visibility (on-line transaction reporting
payment, and other transaction types, such as ticketing. New
services, self service, dispute reconciliation), reduced settle-
transaction types include health insurance payments, P2P
ment times, and improved financial services, examining the
payments, loyalty schemes, mass transit systems ticket pur-
broader business needs, such as cash flow requirements,
chase, etc. Payments processors need to evolve to support
supplier chain integration, insurance, leasing etc.
multi-currency and channel integration capabilities, and more than just standard financial transactions.
Other significant market changes impacting the status quo
Communications — ATM and POS networks have been domi-
Whilst merchant pressure is having a real impact on the pay-
nated by proprietary, costly communications infrastructure.
77
Transactions are typically aggregated by Telcos and forward-
tolerance.
ed on to payments processors. With the advent of ubiquitous Internet Protocol (IP) the potential is for ‘always on’ devices to
A wave of second generation solutions has provided software
route transactions to preferred processors. The legacy aggre-
on open platforms (UNIX, NT etc.). However, to provide the
gation services are threatened and payments processors will
levels of fault tolerance associated with the mainframes/high
be driven towards differentiation beyond terminal connectivi-
end servers, clustering and replication techniques are
ty and financial transaction processing.
deployed. The limitation of these architectures has included
Payments convergence — payments processing has typically
ly complex recovery issues in the event of a partial system
been managed through lines of business. However, an enter-
failure. Whilst these solutions are more flexible and open,
prise view of payments could lead to greater efficiencies and
there is still a heavy reliance on the original software vendor
the production of new products/ services. The same environ-
to customize the software.
reduced performance and scalability dimensions and extreme-
ment could support B2B, P2P, bill payments, high value/lower volume payments, high volume/lower value payments, sup-
The next generation of payments solutions now offer a truly
plier payments, etc.
open and adaptable environment, unlimited scalability, and, importantly, fault tolerance at the application level through a
The technology challenge
distributed architecture. Costs are driven out through choice
The complexity and scale of the payments switching process
of developer/integrator, use of market standard servers and
cannot be underestimated. For example a large payments
databases, speed to market for new services, lower upgrade
processor may handle millions of real-time authorization
costs, and utility models for license fees and infrastructure.
transactions a day from in excess of 100,000 POS/ ATMs
Transaction information is available in real time for distribu-
devices plus internet, mobile, and voucher channels. It has to
tion to merchants, banks, and senior management via web
settle funds with card issuing institutions, card schemes, and
interfaces. The next generation application delivers a signifi-
tens of thousands of merchants. The costs of managing dis-
cantly lower cost base, greater business agility, and actually
pute processes are significant. In addition, the systems have
improved reliability through enhanced replication and moni-
to provide 24 x 7 operations to allow consumers to transact
toring.
anywhere at anytime.
Conclusion However, to meet the demands described above, processors
In delivering a secure, trusted payments network, payments
need to examine their existing infrastructure and assess the
participants have focused on network centric issues to expand
migration risks of moving to new business models, a lower
and grow their businesses. The new dimension is to provide
cost model, and new communications, software, and plat-
client-centric services; value for money, faster settlement
forms.
processes, value-added services, new channels, and greater convenience and transparency.
Large payments processors have had very little choice in
78 - The
selecting the necessary high performance transaction
The costs and risks associated with changing to new business
processing environments needed to drive large payments
models and approaches, coupled with the potential cannibal-
networks. Typically, the solutions have utilized in-house or
ization of existing revenue streams, have hindered innovation
heavily customized vendor software operating on either
and responsiveness. The challenge for existing payments pro-
mainframes or high-end servers for high availability and fault
cessors is to either reduce their costs and adapt or be subject
Journal of financial transformation
to a gradual erosion of their business. However, the real opportunity is for large processors and new entrants to exploit the market changes by providing consolidated services that do not disintermediate the existing network operators. This may be achieved by providing differentiated services and revenue sharing models to banks and merchants within the network on the same infrastructure. The ultimate vision is to deliver electronic payments as a utility service across multiple channels at a national, regional, and potentially a global level. A competitive dynamic is maintained by leveraging the infrastructure to support value added aspects for particular network relationships. In defining the steps to deliver on new strategies, technology has to be a high consideration. The ability to differentiate service, adapt, provide scale, and deliver the highest level of availability are complex dimensions to orchestrate. Importantly, risks may be managed through a phased introduction of next generation technology. The solution needs to meet existing functional and service level requirements as well as realizing the core benefits; reduced costs, increased business agility, and improved service levels. The payments processor’s challenge is identifying what is truly open, what will deliver today’s requirements, and what will rapidly adapt to meet an ever changing and dynamic market.
79
Transition
Simulation: A powerful research tool in payment and settlement systems Harry Leinonen Adviser to the Board, Financial Markets, Bank of Finland1
Kimmo Soram채ki Policy Expert, Directorate General Payment Systems and Market Infrastructures, European Central Bank
Abstract Payment systems are complicated and contain many internal and external dependencies. Society is dependent on well functioning payment infrastructures and its stability and efficiency both require continuous monitoring. It is difficult, if not impossible, to test different kind of risk scenarios and policy or structural changes in daily production systems. Simulation models, payment system simulators, provide a powerful research tool for payment and settlement system analysis. This article describes the general structures and dimensions of payment systems and the possibilities of a payment system simulation approach.
1 The views expressed are those of the authors and do not necessarily reflect the views of the Bank of Finland or the European Central Bank.
81
Simulation: A powerful research tool in payment and settlement systems
Simulation — the possibilities
of the day, a net position is calculated from payment flows for
The development of payment and settlement systems often
each participating bank. In such deferred net settlement sys-
involves the hands-on work of designing and operating sys-
tems (DNS), covering funds are usually transferred, with a lag,
tems. The analytical work must tackle the problem of handling
across accounts at central banks. In order to speed up pay-
of large amounts of data and transactions. Simulation tech-
ment transfers, especially for large payments, online continu-
niques enable the construction of models that closely mimic
ous settlement systems (CNS) have been developed. But
the real world. The generic econometric models are too limit-
because a netting system can give rise to large inter-bank
ed to deal in sufficient detail with the many dimensions of
credit risk positions, authorities have generally required that
payment systems, which entail important dependencies, both
these risks be limited or have provided in their stead real-time
within and between systems. Decision-making and other
gross settlement systems (RTGS).
behavior issues are heuristic and dependent on counterparty and situation. For this reason, payment system models must
Payment and settlement systems are often hierarchically
allow for the embedding of different kinds of behavior models.
structured. The top level comprises central banks’ RTGS systems, where large payment transfers are executed, inter-bank
A reality-friendly simulation model provides a veritable labo-
debits and credits are settled, and other systems’ final trans-
ratory setting, in which one can analyze the probable affects
fers of covering funds are affected. For example, the Bank of
of different structural options and decision parameters on
Finland’s RTGS system (BoF-RTGS) handles inter-bank trans-
payment flows and system participants. An on-stream pay-
fers of covering funds for certain other Finnish systems: Retail
ment system is so sensitive that one cannot use it for direct
payment system (PMJ), express and check transfer system
experimentation without endangering daily operations.
(POPS), and two securities settlement systems (OM for shares and RM for debt instruments). The international dimension of
Another area of research is the study of different crisis sce-
the BoF-RTGS is represented by its continuous linkage with
narios and how to prepare for them. While realizations of such
the European System of Central Banks’ TARGET system,
crisis scenarios are extremely rare, simulation models enable
through which covering funds are transferred for systems
one to study numerous crisis-like situations and thus to pre-
handling private large-value payments in Europe (EURO 1) and international currency transactions (CLS). Figure 12 shows the
pare in advance for the real event.
hierarchical structure of payment and settlement system netNonetheless, simulation models are subject to their own
works in general.
important limitations. Optimization analysis requires complete enumeration and iteration, so that one may actually
Risk considerations
miss the true optimum. There may be problems with the
Large inter-bank risk positions can arise in a netting system
behavioral models and assumptions, which may not hold in
because banks accept incoming payments on behalf of cus-
certain special situations. It is fairly simple to use payment-
tomers before covering funds are transferred between banks.
flow time series in simulation studies, but this may result in
G10 central banks’ Committee on Payment and Settlement
the failure to detect the impact of future changes in payment
Systems, in its Lamfalussy report [BIS (1990)], focused atten-
flows.
tion on this type of risk. Authorities then proceeded to
Basic structures of payment systems
lateral requirements for risk control purposes. The report also
Traditionally, payment systems have operated on the basis of
promoted the development of RTGS systems by central banks
batch-processing and system-specific timetables. At the end
[BIS (1997)]. And in order to reduce risks, core principles have
demand that private netting systems employ limits and col-
82 - The
Journal of financial transformation
Simulation: A powerful research tool in payment and settlement systems
The last several years have seen a growing interest in research
Although the simulator was not intended for outside use, other
and development in inter-bank payment and settlement sys-
central banks made use of it with the help of Bank of Finland.
tems. Central banks have allocated effort and resources to this special area. Among the factors behind the movement are
Experience with the system prompted the Bank of Finland to
international cooperation, technological advancement, grow-
proceed with the development of a new and more diversified
ing dependency on smooth functioning payment systems, and
simulator, designed especially for external use and interna-
risk perspectives. Cooperation has been a particularly impor-
tional distribution. The new simulator, BoF-PSS2, was com-
tant factor in the Euro-zone area, which has seen extensive
pleted in Spring 2004 and is available for research purposes
linking of payment systems in order to give the region a
free of charge.
totally integrated inter-bank payment system. The key official fora have been payment system committees of G10 central
The purpose of this article is to introduce the basic issues and
banks and of the European System of Central Banks, and
research topics related to payment systems that have been studied with the aid of simulation models as well as the pos-
working groups of the European Commission.
sibilities for using the Bank of Finland simulator. A more At the Bank of Finland, simulation methods have been used for
detailed description of payment system features and interest-
several years in the analysis of payment systems. This type of
ing topics for simulation studies can be found in Leinonen and
research was initiated around the time Finland was joining the
Soram채ki (2003).
Economic and Monetary Union and it became necessary to examine the impact that the new methods of transferring payments and covering funds would have on Finnish payment systems. Out of this concern evolved the construction of the first payment system simulator. The simulator proved to be an excellent tool for studying liquidity needs and system risks.
DNS level
CNS level
RTGS level
Domestic currency
RTGS 1
RTGS j
Foreign currencies
Bonds
Shares
Derivatives
RTGS k
RTGS m bonds
RTGS n shares
RTGS p derivatives
PVP-condition
Intraday liquidity injections
DVP-conditions
CNS b bonds
End-of-period RTGS settlement
DVP-conditions
CNS b
End-of-period RTGS settlement
DNS r
DNS s
End-of-period RTGS settlement DNS u bonds
DNS v shares
DNS w derivatives
Figure 1: Hierarchical structure of payment and settlement systems
2 Figure 1 illustrates the linkages for simple and conditional funds transfers. In connection with securities settlement systems, delivery versus payment (DVP) means that delivery of securities is conditional on simultaneous payment. Payment versus payment (PVP) refers to currency trading in which payment in one currency occurs only simultaneously with corresponding payment in the other currency. DVP and PVP were developed as means of reducing settlement risk.
83
Simulation: A powerful research tool in payment and settlement systems
settlement process. Certain timing methods can also be used
ficient money and securities).
for these purposes. In Figure 2 the dotted line shows the possible benefits of a settlement algorithm that solves gridlock
Cost elements
situations in the system. In such situations, the parties’ trans-
Settlement systems entail three structural cost factors that
actions are queued and cannot be settled individually one-by-
are partly interdependent: Liquidity, delay, and credit-risk
one but could be settled by netting or splitting [Leinonen and
costs. Liquidity, such as holdings of funds for completing set-
Soramäki (1999)]. A typical example is a circular situation in
tlements, involves certain costs. These are, aside from storage
which participant A is paying participant B, who is paying par-
costs, mostly opportunity costs, since the interest rate on eli-
ticipant C, who is paying participant A; but no one has suffi-
gible collateral is usually lower than on other investments, due
cient liquidity to pay and thus break the circle.
to the liquidity premium. If settlements and the underlying payments are late, delay costs arise, mostly in the form of
In seeking a partial netting solution, one encounters the
sanctions. If the payment system operates on the basis of
‘knapsack’ problem. This arises when liquidity is not sufficient to complete all the transactions in a queue and one must com-
credit, users also incur credit risk costs. The basic structural interdependencies of cost elements are depicted in Figure 34.
pose a smaller group of transactions for settlement. In seek-
Besides structural costs, every system encounters processing
ing an optimal combination of transactions to settle, one is
costs. These are, however, generally of the same magnitude
faced with so many possibilities that going through the whole
regardless of structural option and depend mainly on the effi-
list is an inefficient procedure. Algorithms are publicly avail-
ciency of underlying computer systems.
able that find the optimal solution if the order of payments is fixed [Bech and Soramäki (2001, 2002)] and an approximately
Credit risk costs do not arise if the system operates com-
optimal solution if payments can be settled in any order from
pletely on the basis of liquidity. If settlement is wholly or
the queues [Guentzer et al. (1998)]. The problem of gridlock is
partly based on implicit credit relations between the partici-
particularly acute in securities settlement systems that
pants, the cost of liquidity will be smaller. The difference,
require delivery versus payment (participants must have suf-
however, is that liquidity costs fall on sending parties while credit risk costs are reduced for receiving parties. Similarly,
Lower bound
4
Upper bound
2
liquidity costs can be reduced by delaying outgoing payments so that they are in synchrony with incoming payments and thus benefit from the netting effect.
Payment delays
Delays in settlement and delivery are usually caused by limits on available liquidity and on permitted credit risks. In practice, delays are greater, the less the liquidity available and the 3
tighter the control of credit risks. In their role as overseers, central banks have usually attempted to increase liquidity in order to ensure adherence to payment processing timetables. At the same time, they have required reductions and compensation via liquidity transfers, especially to avoid large risk Liquidity available
1
positions. Figure 3 gives an example of a system (A) based on full liquidity and a system (B) based on credit.
Figure 2: Relation between liquidity and settlement speed
84
3 The relationship between liquidity needs and settlement speed is studied in Koponen and Soramäki (1998) and Leinonen and Soramäki (1999).
Simulation: A powerful research tool in payment and settlement systems
been drawn up for systemically important payment systems
Liquidity needs
[BIS (2001)].
Liquidity needs of payment systems and participants vary In some payment systems, counterparty risks can be so large
according to timetables and system features. Figure 2 gives a
as to give rise to the threat of a domino effect. Financial prob-
general picture of liquidity needs. The lower bound applies to
lems of one participant can be transmitted via payment sys-
the end of the period, i.e., it is the minimum amount of liquid-
tems to other participants so as to create a systemic risk to
ity needed to settle all payments by the end of the day. If the
the whole financial sector. Integration of payment systems has
final inter-bank settlement and transfer of cover are not car-
increased the likelihood of contagion of systemic risk situa-
ried out until the end of the period, it is sufficient that each
tions across wide areas of the globe. Authorities have
counterparty has settlement funds available in the amount of
attempted to limit the realization probability of such risk by,
its net liability. Settlement in this case is delayed for the
for example, requiring gross transfers of covering funds and
maximum time (point 2). In case payments are settled con-
collateral coverage of all significant counterparty positions.
tinuously with this same amount of liquidity and queued
Systemic risk has also become an important research topic
whenever not enough liquidity is available, the settlement
[Angelini et al. (1995), Kuussaari (1996), and Bech et al.
delays can be reduced (point 3).
(2002)]. The upper bound (point 1) applies to continuous real-time Liquidity risk is related to possible shortfalls of funds for set-
settleÂment. Payments are settled immediately and liquidity
tling payment obligations. Banks do attempt to forecast their
need is maximized. Any liquidity in excess of this amount will
coming liquidity needs, but payment flows are subject to diffi-
not be used. Between these extreme situations is a continuum
cult-to-forecast stochastic fluctuations. These fluctuations
of options for which settlement is delayed and payments are
may be troublesome, especially in times of market distur-
variously queued. The shape of the curve depends on the
bance or when other banks in the system are facing liquidity
extent of the continuous netting effect. If liquidity is reduced
problems. Realization of liquidity risks may even cause a tem-
below point 3 some payments will necessarily remain unset-
porary disruption to a large part of the payments flows.
tled while others are increasingly delayed. If liquidity is reduced to 0 (point 4) no payment can be settled by gross settlement.
Settlement-delay costs
Interest in intraday liquidity derives from developments in payment systems and shrinking delivery times. Payment systems are currently in a kind of interim stage. Earlier, before System A
System B
the 1990s, operations were strictly on the daily level and intraLiquidity-usage costs
day liquidity had no significance. As the processing of payments has been speeded up and central banks have converted to RTGS, intraday liquidity has received increasing emphasis. As speed becomes more important, banks’ liquidity needs increase toward the upper bound3.
Credit-risk costs Figure 3: Structural costs of payment and settlement systems
4 Figure 3 depicts a three-dimensional cost minimization problem in terms of credit risks, liquidity, and payment delay costs. All of the cost dimensions in the chart can be traded off against each other. The relationships with a given technology are usually convex resulting in a unique cost optimum.
If payments are queued, various netting and splitting algorithms can be used to reduce liquidity needs and speed up the
85
Simulation: A powerful research tool in payment and settlement systems
Bank of Finland simulator
Using the simulator
The Bank of Finland Payment and Settlement Simulator can
The Bank of Finland Simulator is a modern research tool that
be characterized as a deterministic system with decision-
has been well received, particularly by other central banks.
making capability, which takes stochastic data as input. For a
Currently, about 20 central banks around the world are using
given input data and choice of algorithms and parameters, the
it. These include all the Nordic central banks, European
simulator will always produce the same results. Usually time
Central Bank, several European national central banks, Bank
series over many days are input in order to get a grasp of the
for International Settlements, Federal Reserve Bank of New
fluctuation ranges of the various results. One can create and
York, Bank of England, Bank of Canada, and several central
use algorithms with decision-making capability that enable
banks in Asia. The simulator is used inter alia for studying
payments processing in accord with different settlement
liquidity policies and counterparty and systemic risks, making
modes and behavior models for settlement parties (banks).
efficiency comparisons, and developing algorithms, pricing
The simulator is built for independent PC usage and is com-
has been particularly helpful in advising on design and in test-
posed of three subsystems — input generation, simulation
ing the new version of the simulator.
policies, and settlement modes. The European Central Bank
execution, and output analysis. Payment and settlement systems have received relatively The simulator operates with Microsoft NT, XP and 2000 and
little attention from financial market researchers. The solu-
requires central memory of at least 256 MB. A users’ guide
tions have usually been based on conventions applied by
and order and installation directions are available at www.bof.
those in charge of actual payment system operations. A
fi/sc/bof-pss. Distribution is handled via Internet. Using Java,
modern automated payment system is based purely on data
the user can build new settlement algorithms and link them to
transfer logistics except for the fact that even a small num-
the given structures. The database is based on the cost-free
ber of bits may contain credit transfers worth billions of
MySQL database.
monetary units. One would hope for an expansion of research in payment systems. The Bank of Finland Payment and Settlement Simulator provides new types of opportunities in this area. Interesting and sparsely-studied areas of research include DVP-based netting and queuing algorithms in securities settlement systems. Related topics of special interest might be models of participant behavior in various stress and emergency situations and the consequences for settlement risks. Various risk management systems that may be incorporated in payment systems represents another area that could benefit from further research (i.e. the reactions of participants and the system to terrorist attacks and to participants’ liquidity problems). The intention is to develop and expand the simulator on the basis of experience and user feedback. It is hoped that an active group of users will share experiences and ideas for
86 - The
Journal of financial transformation
Transition
European payment systems and monetary union
Francisco J. Callado MuĂąoz Assistant Professor, University of Girona
Natalia Utrero GonzĂĄlez Visiting Professor, Universitat Autonoma de Barcelona
Abstract We make a comparative study of payment systems for E.U. -fifteen countries for the 1996-2002 period. Special attention is paid to the introduction of the new European single currency. The overall trend in payments is for a move from cash to non-cash payment instruments, although electronic instruments are not widely used yet. We find a significant impact from the introduction of the new banknotes and coins on card use.
87
European payment systems and monetary union
In a world of economic globalization and IT development the
ments usage in a general setting2 but also to analyze the
ways in which payments are made are clearly evolving. Cash is no longer the unique possibility of making a payment and it
impact of the introduction of the new currency, both among the first 11 countries in 19993 and the full introduction of Euro
shares the stage with credit and debit cards, direct debits, and
notes and coins in 2002 across E.U.-15, on these systems.
electronic means.
Cash use Bankers and other professionals1 are very interested in know-
To analyze the importance of cash payments we use two of
ing how customers are paying for their daily transactions and
the typical proxies used by the BIS studies: cash in circulation
monitoring what changes are taking place. They are not alone, financial authorities are also interested in this type of infor-
as a percentage of GDP and cash in circulation as a percentage of narrow money4. Figure 1 offers these two measures for
mation, since one of their responsibilities is the promotion of
the countries included in the sample. As can be seen, there is
efficiency and security of both payment systems and instru-
a decline in cash use for both indicators and for almost every
ments in order to safeguard the monetary policy transmission
country. The only exception is the U.K., where there is a slight
mechanism and to contribute to the maintenance of systemic
increase in cash usage for the period 1996-2001 in the propor-
stability and public confidence in the currency [ECB (2002)].
tion of narrow money and around 10% for 1996-2002 in the proportion of the GDP. With respect to the distinction between euro and non-euro countries a point has to be made. The euro-
Cash % of narrow money Cash % of GDP
zone shows a decrease for the period 1996-2001 and 1996-
1996
2001
2002
1996
2001
2002
Belgium Denmark Germany Greece Spain France Ireland Italy Luxembourg Netherlands Austria Portugal Finland Sweden U.K. E.U. Euro-zone
27.5 10.2 27.6 44.1 25 14 34 16.1 14.6 18 22.9 15.3 6.6 4.9 13.8 18.1
11.8 9.2 11.3 30.9 12 7.4 16.5 11.3 0.8 5.7 13.9 8.7 6.1 5 8.6 10.3
- 8.7 - - - - - - - - - - - - 4.8 11.5 14.1
5.1 3 6.9 6.5 10.3 3.3 4.7 5.3 2.9 5.3 5.8 4.9 2.2 4.1 3 5.3 5.8
2.8 2.9 3.3 5.5 6.6 2 3.3 4.7 1.9 2.1 3.9 3.6 1.9 4.5 3.3 3.5 3.5
2.9 4.1 3.3 3.5 4.8
Figure 1
2002, but it presents an increase in cash use with the introduction of euro coins and notes in 2002, that is around 35%. Non-euro countries, on the contrary, decrease their use of cash and the figures for 2002 are quite similar to 2001. It can, therefore, be concluded that the usage of cash declined during the period of study, with the exception of the physical introduction of the euro in 2002 that made the euro area more dependent on cash. The ‘dual circulation period’5 of the new euro notes and coins and the former national currencies together with the flourishing of money from the unofficial economy could explain this break in the tendency of cash use reduction of the previous years.
Retail payments Another way of looking at payments systems would be to
88
The purpose of this paper is to study the main trends in pay-
analyze the competition among instruments at retail level. On
ment systems across the E.U.-15 countries for the period 1996-
the one hand, ATM networks allow customers to have access
2002. The data used in our analysis are derived from the
to cash closer to the point of sale. On the other, EFTPOS
European Central bank (data on payments) and the European
instruments, such as debit cards, provide consumers with non-
commission (economic data). This analysis is interesting not
cash means of payment right at the point of sale. Figures 2 to
only to have a portrait of the evolution of payment instru-
5 help understanding their evolution in the last years.
1 Companies different from banks (namely retail stores, insurance, petrol stations etc.) are now offering payment services to their clients mainly trough credit and debit cards. 2 Markose and Loke (2000) made a first attempt to analyze payment usage for the period 1990-1998 and served as a starting point for this work.
3 The introduction of the single currency in Greece took place in 2001. 4 Usually related to M1, although the correspondence is not always exact. 5 The speed of the changeover was not the same in all countries. The dual circulation period lasted between four weeks and two months.
European payment systems and monetary union
dominance of card in value terms. This does not hold for the
High usage of electronic payments
Euro-zone since the six countries in which cash is still the
110
Fi G Fi D
100
90
EU UK
80
most important payment instrument are within the monetary
A
this euro-zone.
G B S EU
With regards to the intensity of card and cash per EFTPOS and
e
ATM terminals in terms of value, although no general trend
Ir
Fr
can be identified in the relative importance of each there is an
It
UK
60
union. Again, in 2002, the ratio indicates a rise in cash use in
A
S D e It
P
70
B
increase in the intensity of card use, that is, each EFTPOS is processing a greater value of transactions. This growth pat-
P
50
Fr
tern holds for almost every country, including the E.U. and
40
Euro-zone countries. Interestingly, for the Euro-zone countries
Ir High usage of cash
30 0
10
20
30
40
this increase also holds 2002. The intensity of cash use by ATM, on the contrary, does not follow a clear tendency. Some countries grow, some others
Figure 2: Cash use and non-cash electronic payments
decline, and some remain nearly constant and figures are more stable in all cases. The E.U. and the Euro-zone experi-
When looking at the relative network densities of EFTPOS to
ence relative growth until 2002, when the increase becomes
ATM terminals we find that there is no regular pattern across
more profound.
all countries. Greece, Italy, Austria, Sweden, and Portugal seem to have an increase in this period although the rate of
It can be concluded then that cards are being more inten-
growth differs across the countries. Germany, Ireland, and
sively used than cash. Consequently, financial authorities and
Spain grow at the beginning of the period and then decrease
bankers could increase more than proportionally the accep-
in the later years, while France, Finland, and Luxembourg are
tance and use of cards, and in the process decrease the use of
relatively unchanged. Denmark, Belgium, and the U.K. do not
cash, just by increasing the relative number of EFTPOS to
show a clear tendency. With respect to the Euro and E.U. coun-
ATMs.
tries, both groups increase at a similar rate although, again, with the introduction of the notes and coins (2002) the euro-
Cash and non-cash electronic payments
zone suffers a slight decrease. This fact follows the same pat-
We also looked at the relationship between cash use (as a
tern as those observed in Figure 1.
percentage of narrow money) and non-cash electronic payments and their evolution from 1996 to 2002 across these
Comparing the per capita ratio of value of card transactions
countries (Figure 2). We found that at the beginning of the
(both credit and debit) to the per capita value of ATM related
period almost every country had a high cash use and a more
use, we find that more than half of the countries have a ratio
variable electronic component. By 2002, most countries had
greater than 1, that is, the value of card transactions is bigger
lower usage of cash and increased the proportion of non-cash
than the value of cash. Remarkable cases are those of France
electronic payments. Only Ireland remains with a high degree
and Sweden that start the period below 1 and end up with
of cash usage, however given its starting point of very high
cards over cash in 2002. The E.U. as a whole presents also a
cash use and low usage of non-cash electronic payments its
89
European payment systems and monetary union
tries and suggest that financial structure, as well as the intro-
withdraw money and to pay, such that customers have access
duction of the single currency, may have affected the decision
to a direct and personal service easily (substitute effect). On
to use them. In this section, we want to investigate more
the other, the closer service associated to a larger number of
deeply the empirical evidence on these relationships. In par-
branches can transmit trust and reliability in the banking insti-
ticular, we test three hypotheses:
tution and foster the use of cards (reliability effect). We will try to shed some light as to which effect is stronger.
■ Whether the facilities developed by financial institutions
The physical facilities, namely number of ATMs and the num-
significantly influence the use of cards instruments?
ber of EFTPOS positively influence the use of cards. The larger
■ If the degree of economic development can also affect
number of ATMs (EFTPOS) the larger the probability of using
card use?
them. Therefore, the expected sign is positive.
■ Whether the introduction of new coins and bank notes
influence the evidenced increasing trend in card use? To capture the effect of the introduction of the euro, we Specifically, we distinguish between the use of cards to with-
include two different dummy variables, one for the year 1999
draw money from bank accounts and the use of cards as pay-
when the European single currency was introduced and one
ment instrument at point of sale. We estimate the following
for the year 2002 when the bank notes and coins were physi-
equations:
cally introduced. Finally, we include the per capita GDP in each country to control for the degree of economic development.
Card usei,t = α + β * banking industry facilitiesi,t + γ*gdpi,t
The expected sign is positive. The use of cards requires a
+ ψt + εi,t
minimum degree of electronic facilities and communications
(1)
to operate correctly. We assume that economic development where i = 1,…, n refers to countries and t = 1,…, T to time periods.
is a good proxy for technological development.
From an econometric point of view, for the estimation of the
Figure 3 presents the results for the withdrawal functions of
coefficients α, β, and γ we take into account the structure of the error terms, ε 8. We also allow for the presence of unob-
cards. We estimate the equation using three different mea-
it
sures of the dependent variable. The cash use of cards is
servable individual effects9.
computed in terms of the total volume of transactions, num-
The variables that account for the banking facilities are the
tion in Panels A, B, and C respectively.
ber of transactions done, and the average value per transacnumber of branches and the number of ATM and EFTPOS available. All variables are controlled for the population of the
90 - The
ATM network influences significantly and positively card use
country in order to allow comparisons. The number of branch-
to withdraw money both in terms of volume and the number
es is included in all regressions. However, we use the number
of transactions, as it can be observed in all runs in Panels A
of ATMs when the dependent variable is the use of card to
and B. However, the average value of transactions is not
withdraw money and the number of EFTPOS instead when the
affected by ATM network (panel C). Column 1 in each panel
dependent variable is the payment function of credit/debit
presents the results when the dummy variable that accounts
cards. The number of branches accounts for the proximity of
for the introduction of the single currency is introduced. The
the banking institution to customers. The expected sign of this
1999 variable is not significant in any of the realizations, that
variable in the use of cards is not obvious ex ante. On the one
is in volume, number of transactions, or average value.
hand, close service to customer might reduce card use both to
However, the dummy for the introduction of the bank notes
Journal of financial transformation
6 Ireland is the country where cheques have the most relevant use both in volume and value, but we find that its importance also falls during this period. 7 It is important to remark that ATM and EFTPOS terminals increase 54% and 86% respectively in the E.U. during the period (computed from Eurostat figures).
European payment systems and monetary union
development is one of the most remarkable of any country.
We also examined the potential impact of the euro on the
We do not find a clear distinction between the E.U. and Euro-
usage of non-cash electronic payments and instruments, by
zone countries. Although the origin is different, less cash use
looking at the changes in the years 1998-1999 and 2001-2002,
in E.U. countries, they both end up in a pretty much similar
and found no ‘euro effect’. We found that there was no break
position by 2002. Therefore, in terms of volume, we find that
in tendency present. On the contrary, results confirm the
there is a clear trend towards lower cash usage and to greater
trends and patterns found in the whole period, in particular
usage of non-cash electronic means of payment.
relative to the value of transactions. Therefore, the introduction of the new bank notes and coins do not seem to have
Evolution of non-cash payment instruments
dramatically impacted the relative importance of non-cash
As shown in Figure 2, there is a trend from cash to non-cash
instruments.
payment instruments, although paper-based instruments are still important in many E.U. countries. What we intend to do in
Econometric analysis
this section is to assess the relative importance of the non-
The previous sections show that both the relative importance
cash payment instruments in volume and value between 1996
of cash and non-cash use and the distribution of non-cash
and 2002.
instruments are not homogeneous across the European coun-
We find that the use of cheques decrease significantly between these two periods6. The use of credit and debit cards increase considerably during this period, although their relative importance in terms of value is small. Hence, it seems that card use is probably more related to daily operations7. Credit transfers account for the majority of total value transactions and are important in terms of volume of transactions, although it falls during the period. Direct debits, instead,
Panel A: Total volume of transactions The dependent variable, Cash value, is the total volume of card transactions to withdraw money. Financial facilities are number of ATMs (Nº ATMs) and number of branch offices (branch) per 1000 inhabitants. D1999 and D2002 account for the year of introduction of the common currency and notes and coins respectively. (1) (2) (3) (4) Cash value Cash value Cash value Cash value Nº ATMs x 1000 inhab
2679939c [428515]
2748867.65c [438,145.177]
2779121.05c [436,816.852]
D1999
-63003.57 [132541.4]
D2002
512,332.699c
[133,264.997]
543,062.722c [134,744.668]
443,695.083c [119,858.977]
period, we find their usage remains quite small, particularly in
Branch
734,546.0198
1788831.82c
value terms. Belgium, Luxembourg, and The Netherlands are
x 1000 inhab
[562,480.867] [499,017.945]
the countries with highest usage of electronic instruments
Per Capita GDP
96,257.126c
[15,444.7848]
and experience the largest increase. This lack of growth takes
Constant
place at a time when access to new technologies and the
694017.1b [314588.3]
575,140.862b [235,788.986]
207,085.4122 [366,794.189]
-1.6117e+06c [443,397.185]
Internet has experienced a huge increase in E.U. countries
Obs
95
95
95
90
(876% in the period 1996-2001 according to Eurostat figures).
R-squared
0.4872
0.57
0.58
0.69
It seems logical, therefore, that should financial institutions
Hausman test
0.00
187.85c
32.92c
125.59c
remain fairly stable in all countries. Although one would expect that electronic instruments would experience a significant increase in popularity during this
wish to increase the usage of electronic money across Europe, they would need to invest heavily in better security systems,
1236233.59c [431,538.405]
Standard errors in brackets a — significant at 10%; b — significant at 5%; c — significant at 1%
transparency, and publicity in order to increase consumer reliability and the success of these payment instruments.
8 The error term, εi,j,t is identically distributed and uncorrelated across observations and with exogenous variables, but cov (εi,j,t, εi,j,s) may be different from zero if t = s. 9 Individual effects can be treated as fixed or random. The problem is not if effects are fixed or random. The problem is whether the effects are correlated to the observable variables. When correlation is present, conditional inference must be
Figure 3: Use of cards to withdraw money
done (fixed effect estimation) [Arellano and Bover (1990)]. The Hausman test show that individual fixed effects are not correlated with the explicative variables, therefore the random effect estimator is consistent. However, in some specifications the Hausman test rejects the hypothesis that the random effect estimator is consistent, therefore fixed effect estimation is used.
91
European payment systems and monetary union
Panel B: Number of transactions (1) (2) transactions transactions x 1000 inhab x 1000 inhab
and coins is positive and significant, in particular in total volume and the average value of transactions . Therefore, it can (3) transactions x 1000 inhab
(4) transactions x 1000 inhab
be claimed that although the changeover to the euro took place in 1999, it is the introduction of notes and coins that enhances the use of cards at ATMs. Hence, the ‘euro effect’
Nº ATMs x 1000 inhab
24005.42c [2460.575]
21598.94c [2858.337]
21423.62c [2897.677]
14271.28c [2998.799]
D1999
78.20867 [705.4583]
D2002
1271.141 [853.0511]
1375.165 [887.9476]
883.5491 [826.8769]
Branches
1537.869
8420.163c
x 1000 inhab
[3637.473]
[3418.001]
due to the rounding might have affected this result as well. As
takes place in 2002. What is most interesting is that transactions do not increase but the value of each transaction is higher. One possible explanation could be the value of the euro bank notes; larger than the national ones. Moreover, perhaps the increase in prices
Per Capita GDP
513.3503c
[104.8599]
the dummy 1999 is not significant, we introduce the dummy
Constant
9254.231c
[3284.545]
10414.3c [3071.285]
9771.093c [3472.797]
-1007.743 [4014.939]
corresponding to 2002 for the rest of the realizations.
Obs
96
96
96
91
R-squared
0.58
0.59
0.60
0.70
The degree of economic development affects positively the
Hausman test
0.000
0.000
0.000
0.000
cash use of cards, independent of the type of measure introduced (Column 4). Hence, the more developed the country,
Panel C: Average value of transactions The dependent variable, transact (respectively Atmvaptr), is the number of card transactions to withdraw money per 1000 inhabitants (the average value of each transaction in Panel C).
the 2002 time variable (Column 3), its coefficient is not sig-
(1) Atmvaptr
(2) Atmvaptr
(3) Atmvaptr
(4) Atmvaptr
nomic development, its coefficient is positive and significant
Nº ATMs
47.3895c
x 1000 hab
[10.82327]
21.73221b [11.149]
21.1738b [11.252]
1.984646 [12.9219]
D1999
.22863 [3.1227]
the more use of cash instruments. The effects of the branch network are not conclusive. When it is introduced alone with nificant. However, when it is included with the degree of eco(Column 4). In this case, the ‘reliability effect’ overcomes the substitutive effect of larger branch network. The fact that the branch network is only significant jointly with the economic development may mean that the ‘reliability effect’ is taken
D2002
16.36994c
[3.4754]
16.91159c [3.587]
14.5724c [3.8473]
branch x 1000 hab
9.372596 [14.008]
22.6672 [14.5687]
Per Capita GDP
1.28423c
[0.4399]
Constant
80.5646c
[11.9013]
92.616c [10.6675]
88.4251c [12.301]
64.1529c [14.794]
Observations
95
95
95
90
transactions in Panels A, B, and C respectively. In this case, we
R-squared
0.21
0.38
0.38
0.42
replace the number of ATMs with the number of EFTPOS avail-
Hausman tests 2.42
4.24
0.16
3.97
Standard errors in brackets a — significant at 10%; b — significant at 5%; c — significant at 1% Figure 3 (continued): Use of cards to withdraw money
into account by consumers in economically developed environments. Figure 4 presents the results for the use of cards as a payment instrument. Again, we measure the use of cards in terms of total volume, number of transactions, and average value of
able. The EFTPOS network influences significantly and positively card use to pay both in terms of volume and the number of transactions (Panels A and B of Figure 4). However, the average value of transactions is not affected by the EFTPOS network (Panel C).
92 - The
Journal of financial transformation
European payment systems and monetary union
Panel A: Total volume of transactions The dependent variable, Card value (transact respectively in panel B), is the total volume of card transactions to pay (number of transactions per 1000 inhabitants in panel B). Financial facilities are number of EFTPOS (Nº EFTPOS) and number of branch offices (branch) per 1000 inhabitants. (1) (2) (3) (4) Card value Card value Card value Card value Nº EFTPOS x 1000 inhab
156,810.932c [15,586.6287]
121,445.9548c [17,011.3802]
D1999
-42,835.8330 [109,146.4338]
D2002
123,352.5818c [16,356.7001]
78,832.6894c [20,965.4774]
Panel C: Average value of transactions The dependent variable, posvaptr, is the average value of card transactions to pay. Financial facilities are number of EFTPOS (Nº EFTPOS) and number of branch offices (branch) per 1000 inhabitants.
(1) Posvaptr
(2) Posvaptr
(3) posvaptr
(4) Posvaptr
Nº EFTPOS x 1000 inhab
0.0601 [0.1939]
-0.1498 [0.2267]
-0.1514 [0.2242]
-0.7356b [0.2949]
D1999
-0.6359 [1.3386]
D2002
2.6534a
458,201.2561c
400,076.4758c 382,155.4983c
[1.5227]
3.0135b [1.5222]
1.7008 [1.4256]
[115,378.6766]
[112,836.9966] [110,798.7588]
Branch
9.7409
16.9185c
Branch
-1.3803e+06c
x 1000 inhab
[6.0324]
[5.6443]
x 1000 inhab
-977463.0924b [428,152.4659] [410,066.2826]
Per Capita GDP
0.7928c
Per Capita GDP
49,812.2575c
[0.2243]
[15,261.9106]
Constant
53.7210c
32,956.2763
683,449.0571b
[4.2114]
55.3227c [4.2955]
50.6277c [5.2119]
34.9544c [6.3063]
Constant
-243328.4348
[276,408.5673] [278,312.3132]
-220039.2793 [324,643.9854] [404,918.4418]
Observations
105
105
105
100
Obs
105
105
105
100
R-squared
0.01
0.02
0.04
0.27
15
15
15
HausmanTest
0.70
0.28
4.63
3.78
Number of cntry 15 R-sqd
0.24
0.27
0.40
0.53
Hausman test
2.60
1.54
0.03
5.84
Standard errors in brackets a — significant at 10%; b — significant at 5%; c — significant at 1%
Panel B: Number of transactions (1) (2) (3) (4) Nº trans Nº trans Nº trans Nº trans x 1000 inhab x 1000 inhab x 1000 inhab x 1000 inhab Nº EFTPOS 3,056.3032c 2,389.7544c 2,409.6415c 1,779.6621c x 1000 inhab
[300.3363]
[330.6613]
[328.3495]
d1999
-763.6574 [2,085.3649]
[468.6010]
d2002
8,449.3579c
[2,226.9987]
7,673.1950c [2,251.3461]
7,855.6110c [2,376.2326]
Branch
-18,881.4428b
x 1000 inhab
[8,683.1587]
-11,658.9786 [9,080.0094]
Per Capita GDP
726.5331b
[348.2890]
Constant
-3,146.8811 [5,903.6937]
2,095.4014 [5,955.6091]
11,055.6887 [6,824.2131]
-2,927.7474 [9,377.8244]
Observations
101
101
101
96
R-squared
0.23
0.25
0.39
0.40
Hausman Test
1.86
1.19
0.06
4.53
Figure 4: Use of cards to pay
93
European payment systems and monetary union
Conclusion As in Figure 3, Column 1 in each panel presents the results
This paper illustrates that the use of cash declined between
when the 1999 dummy is introduced. The 1999 variable is not
1996 and 2002. There was an increase in non-cash payment
significant in any of the realizations. However, the 2002
instruments due to new technologies, in particular card use.
dummy, which accounts for the introduction of the bank notes
Accordingly, traditional instruments such as checks experi-
and coins, is positive and significant. Therefore, it can be
ence a decline, but surprisingly electronic money relevance is
claimed that there is a euro effect that also enhances the use
still very small. The econometric estimations confirm the
of cards as payment instrument. A possible interpretation
descriptive analysis results concerning cash use. Furthermore,
could be that consumers prefer to use cards to pay due to lack
there is evidence that the introduction of the euro banknotes
of knowledge and confidence in the new bank notes.
and coins enhances card use significantly, both at ATMs and EFTPOS. Economic development and financial systems facili-
Contrary to ATMs use, the number of EFTPOS transactions do
ties affect card use as well. Hence, reliability in the financial
increase significantly in 2002, although the value of each
system is crucial for the use of cash less instruments.
transaction is not altered significantly. This result reinforces
Therefore, should financial institutions want to support the
the hypothesis that the incremental increase in value of ATM
use of these payment instruments they would need to invest
transactions is caused by the larger new bank notes, rather
in security systems, transparency, and publicity.
than the increase in prices after the introduction of the single currency. If the latter was the cause it would also have impact-
References
ed the value of EFTPOS operations.
• Arellano, M. and O. Bover, 1990, “La econometría de datos de panel”, Investigaciones Económicas (Segunda Época), 14(1), 3-45. • Eurostat. 2004. “Economic data”. European Commission. http://epp.eurostat.cec. eu.int/portal/page?_pageid=1090,1137397&_dad= portal&_schema=PORTAL • Greene, W. H., 1998, Análisis econométrico. Tercera edición, (Prentice may-Madrid). • Markose S. M. and Y. J. Loke, 2000. “Changing trends in payment systems for selected G10 countries and E.U. countries 1990-1998”. International Correspondent Banking Review , Yearbook, 2000/2001, Euromoney Publication • European Central Bank, 2004. “Blue Book, Payment and securities settlement systems in the European Union, Addendum incorporating 2002 figures”. April • European Central Bank, 2002. “E-payments in Europe the Eurosystem’s Perspective”. September • Bank for International Settlements, 2004, “Statistics on payment and settlement systems in selected countries, figures for 2002,” Committee on Payment and Settlement Systems, March
Branch network has a significantly negative impact on use of cards for payment, in terms of volume and number of transactions. Therefore, in the case of EFTPOS, it seems that the substitute effect is dominant. Results for the average value of transactions, however, are not conclusive. When the branch variable is introduced alone with the 2002 time variable, its coefficient is not significant (Panel C); but, when it is included with the degree of economic development, its coefficient is positive and significant. In this case, the reliability effect overcomes the substitutive effect of larger branch network. Therefore, the trust in the financial sector is important for the increase in the value of card operations. Finally, we introduce the degree of economic development. The coefficient is positive and significant in all three cases, as expected. Hence, the more economically developed, the more use and trust in cards, both to withdraw and to pay.
94 - The
Journal of financial transformation
Transition
Karen Furst Policy Analyst, Office of the Comptroller of the Currency
Technological innovation in retail payments: Key developments and implications for banks1
Daniel E. Nolle Senior Financial Economist, Office of the Comptroller of the Currency
Abstract The United States still has a heavily paper-based retail payment system when compared with many other developed economies, but the shift to electronic payments has been bigger and more decisive than commonly perceived. For the first time ever, check use declined in the mid-1990s, and among electronic payments the adoption of debit cards has occurred at an extremely rapid pace. The purpose of this article is to promote greater awareness of the nature of recent changes in retail payments in the United States, and to explore some of the key implications of those changes for the banking industry. We describe recent trends in retail payments, highlighting the surprising drop in check usage, and consider the impact of changes in retail payments on bank revenue and costs. We conclude that, although banks are likely to realize substantial long-run payments-related improvements in productivity, they face greater-than-commonly-appreciated near-term challenges.
1 The opinions expressed in this article are those of the authors alone, and do not necessarily represent those of the Office of the Comptroller of the Currency or the U.S. Treasury Department. The authors would like to thank David Nebhut, Mark Levonian, and Jeffrey Brown for helpful comments.
95
Technological innovation in retail payments: Key developments and implications for banks
Recent changes in the retail payment landscape
annual totals for each year. From this it is clear that both the
The United States has long been considered unusual among
magnitude of consumer and business reliance on checks, and
developed economies for its reliance on paper checks, and
the trend in that reliance, had been off by a factor of 25 per-
hence for its proportionately lower use of electronic means of payments.4 The annual publication by the Bank for Interna
cent or more.
tional Settlements (BIS) of payment system statistics for the
The improvement in the accuracy of the check usage data has
G-10 countries in its ‘Red Book,’ considered to be the most
direct implications for our view of electronic payments. These
authoritative source for such information, ratified the percep-
are clearly illustrated in Figures 2 and 3, which focus on the
tion of continuing, though slowing, growth of check usage in the United States.5 On the basis of such information, payment
pre-2000 period. Figure 2 shows that there was a greater reliance on electronic payments than had been previously sup-
system officials, practitioners, and analysts commonly
posed. In particular, whereas the old (with original check data)
observed that, although U.S. consumers and businesses were
perception was that in the late 1990s somewhat more than a
steadily increasing the use of electronic means of payment,
quarter of all non-cash retail payments were undertaken with
they nevertheless continued to cling to check usage.
an electronic medium, the true figure (with revised check data) was closer to 40 percent. Figure 3 illustrates that for the
It was therefore a tremendous surprise when, in 2001, the Fed
first time ever, electronic payments actually exceeded pay-
announced the results of its survey of the use of retail pay-
ments by check beginning in 2002. In that year, 42.3 billion
ments in 2000 [Gerdes and Walton (2002)]. In particular, data
consumer and business payments were conducted via credit
on checks showed that compared to previous survey results
cards, debit cards, or the ACH (automated clearing house)
for 1995, check usage was lower for the first time ever. In sub-
system, compared with 40 billion payments via check. Clearly,
sequent annual editions of the BIS ‘Red Book’ on payment
consumers and businesses have decisively shifted payment
statistics, the Fed restated its estimates of check usage for pre-survey years.6 As Figure 1 illustrates, those restatements
patterns toward electronic media.
show not only a declining trend in check usage, but lower
1991e
1992e
1993e
1994e
1995
1996e
1997
1998
1999
Original data Revised data
57.47 43.50
58.40 44.63
60.30 45.79
61.67 46.98
62.92 49.5
64.68 48.15
66.09 46.57
67.00 45.17
68.00 43.81
Figure 1. Changing perception of check usage in the U.S. (billions of transactions) Source: OCC using data from statistics on payment and settlement systems in selected countries (Red Book), Bank for International Settlements (BIS) (various years). Note: 1996 ‘revised data’ is an estimate consistent with revisions made to Red Book data for 1997-2000 by the BIS. 1991-1994 ‘revised data’ are estimates based on the average annual rate of check use growth between 1979 and 1995, as per Gerdes and Walton (2002).
1991e
1992e
1993e
1994e
1995
1996e
1997
1998
1999
With original check data With revised check data
18.64 23.24
19.87 24.49
20.75 25.63
22.30 27.37
24.11 28.78
25.94 32.06
24.43 31.45
26.76 35.15
29.33 39.18
Figure 2: Changing perception of importance of electronic payments : Electronic payment transactions as a percent of all non-cash payments Source: OCC using data from statistics on payment and settlement systems in selected countries (Red Book), Bank for International Settlements (BIS) (various years). Note: 1996 ‘revised data’ is an estimate consistent with revisions made to Red Book data for 1997-2000 by the BIS. 1991-1994 ‘revised data’ are estimates based on the average annual rate of check use growth between 1979 and 1995, as per Gerdes and Walton (2002).
96
2 Analysts and practitioners divide payments into wholesale and retail payments. Wholesale payments consist of very large-value payments, especially inter-bank payments related to banks’ clearing and settlement role. Retail payments include consumer-to-business and business-to-business payments. The major components of retail payments in the United States include cash, checks, credit cards, debit cards, and automated clearing house (ACH) transactions. Unlike the other forms of retail payments, reliable records for the number and value of cash payments are not maintained, and hence exact data on cash usage is impossible to obtain (and even estimates are generally conceded to be of the ball park variety). For this rea-
son, most analyses of retail payments focus on non-cash payments, a convention we follow in this article. 3 We, and most others, refer to the Federal Reserve’s (2002) Retail Payments Research Project as a single study, although it is in fact a series of three closely related studies conducted by the Fed in conjunction with several consulting firms. The component studies are the ‘Depository Financial Institution Check Study,’ the ‘Check Sample Study,’ and the ‘Electronic Payment Instruments Study.’ Gerdes and Walton (2002) is considered an authoritative source for a description and analysis of the retail payments study.
Technological innovation in retail payments: Key developments and implications for banks
The stunning advancements in telecommunications and infor-
throughout the 1990s (albeit at a progressively slower rate),
mation management in the early-to-mid-1990s profoundly
the new study showed that for the first time ever, check usage
affected the financial services industry. New products, such as
actually declined in the mid-1990s. A necessary corollary to
credit derivatives were developed, access to credit was sig-
this development is that the shift to electronic payments was
nificantly increased as a result of major improvements in
more dramatic than many had imagined.
credit scoring and the development of securitization, and new methods of delivering financial services, especially via the
These profound changes in retail payments trends are likely
Internet, emerged. One aspect of financial services provision
to have major implications for financial intermediaries, par-
that had, prior to this period, remained below most radar screens was retail payments.2 However, beginning in the ear-
system. The purpose of this article is to promote greater
ly-to-mid-1990s, a growing number of practitioners, analysts,
awareness of the nature of recent changes in retail payments
ticularly for banks, which are at the heart of the payment
and policy makers began to focus on the importance of retail
in the United States, and to explore some of the key implica-
payments on financial intermediaries’ performance, competi-
tions of those changes for the banking industry. The next
tive position, and relationships with third-party payments
section describes recent trends in retail payments, highlight-
services vendors.
ing the surprising drop in check usage, followed by an examination of the implications of changes in retail payments on
A key feature of the retail payments landscape is the long-
bank revenue and costs, which concludes that, although
term shift away from paper to electronic means of payments.
banks are likely to realize substantial long-run payments-
Recently, a comprehensive study of retail payments in the United States was published by the Federal Reserve System.3
related improvements in productivity, they face greater-than-
The study revealed a startling, and hitherto unrealized
tion summarizes and raises a number of payments issues that
commonly-appreciated near-term challenges. The final sec-
change: whereas it had been commonly perceived, and annu-
warrant further consideration by bankers and financial sys-
ally reported by the Fed, that check use continued to grow
tem regulators.
1995
1996e
1997
1998
1999
2000
2001
2002
2003e
Electronic Check
20.0 49.5
22.7 48.0
21.4 46.6
24.5 45.2
28.2 43.8
32.7 42.5
37.6 41.2
42.3 40.0
50.7 38.0
Figure 3. Electronic payments overtake checks (billions of transactions) Source: OCC using data from Statistics on payment and settlement systems in selected countries (Red Book), Bank for International Settlements (BIS) (various years); NACHA; Card Industry Directory 2004; and Minehan (2004).
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003e
Credit card Debit card ACH ATM
4,7 0,22 1,71 6,4
5 0,289 2,49 7,21
5,6 0,71 2,6 7,7
6,6 1,1 2,93 8,45
9,1 1,44 3,49 9,7
10,4 2,3 4,05 10,68
11,5 3,54 4,55 10,98
12,6 4,9 5,33 11,16
14,2 6,49 6,12 10,9
16,9 8,4 6,94 12,84
18,4 10,33 8,06 13,58
19,7 13,22 88,83 13,97
21,1 18,27 10,02 14,27
Figure 4. Growth of electronic payments (billions of transactions) Source: OCC using data from statistics on payment and settlement systems in selected countries (Red Book), Bank for International Settlements (BIS) (various years); NACHA; Card Industry Directory 2004; and EFT Data Book – 2004 Edition.
4 Unless otherwise specified in this article, the terms ‘electronic means of payment’ and ‘electronic payments’ refer to credit card, debit card, and ACH payments. Our comparisons of means of payments focus on the number of transactions, rather than the dollar-value of transactions. For international comparisons of payment systems see Hancock and Humphrey (1998). 5 Statistics on payment and settlement systems in selected countries, Bank for International Settlements. The source for each country’s data is its central bank, which of course is the Federal Reserve in the case of the United States. It is important to understand that in compiling annual payment statistics the Fed uses various estimation techniques, and relies on third-party sources in some instances,
because it is infeasible to get an actual count of all non-cash payments transactions, and impossible to get an actual count of all cash transactions. Note that a thorough survey, such as the Fed undertook for 1979, 1995, and 2000, entails expenses that few would argue are justified on an annual basis. 6 The revisions in the Red Book included data from 1997-2000. To arrive at a revised figure for 1996, we applied the same 3 percent per year estimated decline in check use the Fed used to re-state the 1997-2000 figures. Gerdes and Walton (2002) estimated that, on average, check use grew at a 2.6 percent annual rate between 1979 and 1995, a method we used to generate revised check use figures for 19911994.
97
Technological innovation in retail payments: Key developments and implications for banks
siderable source of revenue. In particular, Radecki (1999)
In addition to the decline in check-related revenue, there are
found that payments constitute between one-third and two-
several other payments-related forces likely to have a nega-
fifths of operating revenue for the twenty-five largest banking
tive impact on payments revenue for banks. Firstly, many of
companies. Furthermore, for banks of all sizes, payment
the electronic payments replacing paper checks are lower
activities are central to deposit taking and lending relation-
margin activities for banks. A recent Ernst & Young report rather gloomily concluded that ‘although many banks have
ships.
spent years thinking about and planning for a decline in paper A large component of payment revenue derives from check
payments, the realization of this phenomenon was still pain
processing and checking account-related activities. Such
ful… Declining paper volumes were seen for [corporate] con-
income streams include revenue from cash management
trolled disbursements, as well as in check clearing, account
services, such as processing check payments at the lockbox,
reconciliation, and in both retail and wholesale lockbox. While
interest income from deposit account balances, and fees associated with insufficient funds.8 Declining check use means
EDI, and purchasing cards, the revenue declines seen from
this revenue is shrinking and will continue to shrink.
in theory there should be an upside for products such as ACH, paper products are not being offset by the replacing electronic alternatives’ [Ernst & Young (2003, p. 4)]. Even among electronic payments, banks are seeing slower growth for those forms of payments, such as credit cards, with relatively
Ratio of transactions 2003 to 1991
higher margins, compared to other forms of electronic pay-
ATM
2.23
Credit card
4.49
ACH
5.87
Secondly, recent changes in fee structures for electronic pay-
81.87
many banks. An important example is the 2003 drop in inter-
ments, such as online debit.9
ments have had a negative impact on payments revenue for Debit card
change fees on signature debit card payments (using Visa check card or debit MasterCard), a decline that came about as a result of shifting market forces.10 The main participants in the market for signature debit card (and credit card) payFigure 5: Proportional growth rate of components of electronic payments: Debit card use explodes 1991-2003 Source: OCC using data from NACHA; Card Industry Directory 2004; and EFT Data Book – 2004 Edition
ments are the merchants receiving the payments; merchant acquirers, which are firms, including some banks, that buy or acquire debit and credit card receipts from merchants, cardissuing banks, and the two main card associations, Visa and
1991
1995
1999
2003e
ATM
49.12
40.84
28.89
22.42
ACH
13.09
14.71
16.22
15.73
Debit cards
1.71
6.08
17.22
28.70
Credit cards
36.08
38.36
37.67
33.14
MasterCard. Merchants sell their card receipts to merchant acquirers at a discount to face value. The size of this discount
Figure 6: Electronic payments: Shift to debit card usage (percent of total electronic payments transactions) Source: OCC using data from NACHA; Card Industry Directory 2004; and EFT Data Book – 2004 Edition.
98 - The
Journal of financial transformation
is largely based on the interchange fee merchant acquirers must pay to card-issuing banks to complete the processing of the card payments. In turn, the two card associations are instrumental in determining interchange fees. A series of recent events brought about a shift in the relative market power of these market participants, resulting in a substantial decline in some interchange fees earned by card-issuing
7 On the banking industry’s long-term prospects for realizing net benefits from innovations in retail payments see, for example, Minehan (2004).
Technological innovation in retail payments: Key developments and implications for banks
ceived shift from paper checks to electronic payments is the
The impact of changing trends in retail payments on the banking industry
change that occurred in the use of various components of
The greater-than-initially-believed switch away from paper
An important feature of the greater-than-previously-per-
electronic payments, such as credit cards, debit cards, and
checks to electronic payments has had, and will continue to
ACH transactions. Figure 4 illustrates trends in credit card
have, a significant impact on the banking industry. Conven
transactions, debit card transactions, ACH transactions and,
tional wisdom holds that advancements in retail payments
for comparison, ATM transactions. In the first half of the 1990s
lower banks’ operational costs, making them better off than
these components of electronic payments rose more or less in
they were in the more paper-intensive era. In fact, however,
tandem. Beginning in the mid-1990s, however, debit cards,
the story is more complex. Although payment system experts
which had been in use since the early 1980s, suddenly surged,
consider long-term prospects for improved bank productivity
overtaking both ACH and ATM transactions.
promising, it is far from clear to what extent banks will realize
Figure 5 puts the growth of debit card use in perspective,
net benefits from changes in retail payment patterns in the near-term.7 Such a conclusion rests on a consideration of
showing that while ATM, credit card, and ACH usage grew
developments on both the revenue and cost sides of banks’
substantially, debit card usage increased more than 80-fold
retail payment activities.
over the 1991 to 2003 period. As a consequence of this change in payment practices, debit cards, which had accounted for
Banks’ payment-related revenue
only 2 percent of all electronic payment transactions in 1991,
Traditionally, banks have focused their efforts on deposit tak-
had risen to 29 percent of all electronic payments by 2003, a
ing and lending activities, regarding payment services in the
position rivaling credit card usage (Figure 6).
same light as other, less glamorous, back office activities. However, for many banks payments activities represent a con-
8 There is some uncertainty associated with insufficient funds fee. Fees for bounced checks and overdraft fees on checking accounts have become an important generator of low-risk income for some banks. Moebs (2003) cites a March 2003 report by the research firm Celent estimating that fees for bounced checks generated between 25 and 50 percent of all deposit fees earned by U.S. banks. Note that recently such fees have come under attack by consumer groups and under scrutiny by regulators. For example, bounced check fees were the subject of a June 1998 report by the Consumer Federation of America entitled ‘Bounced Checks: Billion Dollar Profits II.’ Numerous consumer advocacy groups have expressed their concern about bounce protection programs, including their use as a very expensive form of short-term borrowing. Those concerns are not limited to paper checks. The same groups have urged that financial regulations treat fees charged by banks when an account is overdrawn with an ATM or debit card as an unfair and deceptive practice, on the grounds that no overdraft would have occurred if the bank had declined the transaction. See for example comments submitted by the National Consumer Law Center, Consumer Federation of America, Consumers Union, National Association of Consumer Advocates, and the Woodstock Institute to the Federal Reserve System on the Proposed Amendments to Regulation DD (August 6, 2004). For examples of regulatory responses to the debate, see, e.g., Interagency Guidance on Overdraft Protection Programs, and proposed changes to Regulation DD (which implements the Truth in Savings Act), which were published for comment on June 7, 2004 in the Federal Register (Volume 69, No.109 at 31858 and 31760, respectively). 9 In 2003, the volume of credit card and online debit card payments grew at annual rates of 7 percent and 27 percent respectively. These differing growth rates have a large impact on the revenue of banks that issue the cards, in large part because of differences in ‘interchange fees’ received by the card-issuing banks. The interchange fee is significantly higher for credit cards than for online debit. For example, on a U.S.$75 grocery store purchase, the U.S.$0.91 (VISA) credit card interchange fee would be 6 times greater than the U.S.$0.15 PIN debit interchange fee (specifically, in this case, via the STAR network). The difference in margin increases as the purchase amount increases, because most PIN debit networks cap interchange between U.S.$0.15 to U.S.$0.60 per transaction, while VISA interchange is generally
a percentage of the purchase amount (generally between 1.2% and 2.6%), plus a fixed fee that is usually between U.S.$0.05 to U.S.$0.10 per transaction. 10 There are two ways debit card transactions can be made, either the debit card holder completes a transaction by signing his signature (a so-called ‘signature’ or ‘offline’ transaction), or uses a personal identification number (PIN) at the point of sale (a sometimes referred to as an ‘online’ payment). In general, merchants prefer PIN debit card transactions to signature transactions because interchange fees tend to be significantly lower for PIN debit, ultimately resulting in less of a discount for merchants on the value of the card receipts they sell to merchant acquirers. Note, however, that differences between the two types of debit transactions are narrowing. The Card Industry Directory 2004 reports that interchange fees typically account for two-thirds or more of a retailer’s cost of accepting bank payment cards. 11 The interchange fees market had been in turmoil for some time, as retailers saw a steady increase in the discount rates to which their card receipts were subject. Many retailers believed these exceeded the major costs of issuing cards and processing transactions that the interchange system was originally designed to cover. Indeed, some retailers thought interchange fees should have dropped substantially, as electronic card transactions replaced paper-based card payments. Wal-Mart and other major retailers challenged the pricing power of the two major card associations (Visa and MasterCard) in a class-action lawsuit. The lawsuit challenged the card associations’ ‘Honor All Cards Rule’ requiring retailers accepting one type of card branded by Visa or MasterCard to accept all cards bearing that logo (i.e., if the store accepted branded credit cards they also had to accept the card companies’ branded signature debit card). The parties settled out of court in early 2003; as a result of the settlement, starting in August of 2003 bank issuers of signature debit cards experienced a one-third drop in merchant acquirer-paid interchange fee revenue. Although this decline has been offset somewhat by subsequent increases in interchange fees for signature debit, and by increases on interchange fees for PIN debit card transactions, industry experts estimate that interchange fee revenue has dropped by approximately 20 percent from the pre-settlement level. For a thorough discussion of the pressures on interchange see Giesen (2004).
99
Technological innovation in retail payments: Key developments and implications for banks
The bill payment market is dominated by the CheckFree
new but rapidly growing market. The expedited late payments
Corporation. CheckFree’s electronic commerce division lets
market is led by companies such as Western Union and
customers receive electronic bills over the Internet, pay bills
MoneyGram. Other non-bank providers of so-called ‘walk-in’
electronically, and make payments not related to bills. The
bill payment services have also entered this market, including
majority of CheckFree’s customers access its system through
American Payment Systems, which services up to seven mil-
banks, including Bank of America, Wachovia, Washington
lion households. Note that American Payment Systems was
Mutual, and Wells Fargo. Fees to CheckFree include an average
acquired by CheckFree in June 2004.
U.S.$0.35 charge per transaction for banks that use Check Free for bill payment and e-bill processing, and an average of U.S.$0.81 per transaction charge for banks that fully outsource bill payment to CheckFree [J.P Morgan Securities (2004)]. It is noteworthy that even though banks pay CheckFree such fees, an increasing number of banks do not charge their customers for bill payment. Ebay, the leading online-auction company, purchased PayPal, the leading online payment system, in October 2002. PayPal built its online payment network on the established infrastructure of bank accounts and credit cards. During the second quarter of 2004, PayPal handled U.S.$4.35 billion of gross payment volume (a 53 percent increase over the second quarter of 2003), generating U.S.$161.5 million in transaction fees. For the second quarter of 2004 PayPal’s average transaction revenue rate was 3.64 percent and its processing expenses rate was 1.34 percent, resulting in a 63 percent margin.14 Note that the PayPal system has also begun to be used for crossborder remittances. Expedited late payments are generally for last minute transfers by consumers to pay recurring bills. This is a relatively
2000
Per unit cost — in cents
1.6
ACH volume — billions of transaction 3.81
2001
2002
2003
1.5
1.3
1.1
4.45
4.99
5.59
Figure 7: Fed ACH processing: Per-unit cost declines as volume increases Source: OCC using data from the Federal Reserve System and NACHA.
100
12 See, for example, Porter (2004). His three percent figure is based on information from a Pew Hispanic Center report released in June 2004. Note that several large banks, including Citibank, Bank of America, and Wells Fargo are actively marketing remittance services. In addition, First Data’s 2003 annual report indicates that its Western Union affiliate is encountering expanded competition from banks and ATM providers, which have begun to target money transfer users with new product offerings. 13 Western Union contributed 35 percent (U.S.$2.94 billion in 2003) to First Data’s total revenue (U.S.$8.4 billion) and accounted for approximately 80 percent of the
revenue in First Data’s Payment Service unit, which in 2003 had an operating margin of 34 percent. 14 PayPal has recently completed technological innovations and has expanded its product offerings to become a payments provider outside of the online auction market. In particular, it is revamping its sales force to become a payments provider for retailers. Merchants with U.S.$100,000 or more per month in PayPal volume pay a 1.9 percent discount fee and U.S.$0.30 per transaction. See Internet Retailer (2004) and Bliss (2004).
Technological innovation in retail payments: Key developments and implications for banks
banks.11
Finally, banks are not dominating other higher margin payments activities that have emerged as a result of, or received a boost from, technological advancements. Chief among such payments activities are money transfer (including especially cross-border remittances), bill payment, payments geared to the online market (such as PayPal), and expedited late payments. With the exception of expedited late payments, there is one dominant, and non-bank, provider in each of these increasingly important payments market segments. Money transfer is one of the most important financial services for unbanked consumers and is dominated by non-bank firms, such as Western Union (a part of the First Data Corporation) and MoneyGram. Industry analysts forecast money-transfer growth of 18 percent through 2010, with much of this growth coming from remittances, a payment process that enables immigrants to send money from the United States to their home countries. Some banks have recently shown a somewhat belated interest in the remittance market, but remain small players. For example, banks handle less than 3 percent of the U.S. to Mexico remittance market.12 An indication of the high-margin nature of the money transfer business can be inferred from First Data Corporation’s 2003 performance. First Data reported an operating margin of 33 percent, due in part to a 35 percent contribution to First Data’s total revenue, and an 80 percent contribution to First Data’s Payment Services unit, by its Western Union subsidiary; Western Union has a 75 percent share of the worldwide remittance market, a market in which the average remittance fee is 7 percent of the amount of money transmitted.13
1999
2000 2001
2002 2003
Per unit cost — in cents
3.8
4.0
4.5
4.5
5.1
ACH volume — billions of transaction
17.1
17.0
16.9
16.6
15.8
Figure 8. Fed check processing: Per-unit cost rises as volume declines Source: OCC using data from the Federal Reserve System and NACHA
15 Check electronification can be accomplished via two basic routes. In the case of ‘check conversion’ a paper check is used as a source of information at the beginning of the payment process to initiate an ACH transaction. ‘Check truncation’ is the process of stopping the paper during the clearing process, thus eliminating additional physical handling and transportation of the items. Note that the same economic forces that have propelled check electronification have also stimulated significant improvements in the processing of paper checks. See, for example, Greenspan (2003), and Ferguson (2003).
101
Technological innovation in retail payments: Key developments and implications for banks
added, it is frequently not advisable, or even feasible, to
Conclusions
abruptly discontinue older options, even if those options are
The United States still has a heavily paper-based retail pay-
more costly for a bank [Dove Consulting (2004)]. This lesson
ment system compared with many other developed econ
was most recently learned by banks that successfully inte-
omies, but the shift to electronic payments has been bigger
grated Internet banking into their overall operations. In par-
and more decisive than commonly perceived. For the first
ticular, banks with well-established and well-regarded Internet
time ever, check use declined in the mid-1990s, and among
banking operations analyzed customer behavior, and discov-
electronic payments the adoption of debit cards has occurred
ered that customers adopting Internet banking still wish to
at an extremely rapid pace. These changing patterns of pay-
use the bank’s other delivery channels, including branches and call centers.18 The same dynamic holds true for pay-
ments may eventually result in substantial payments-related cost reductions for banks, but the story on net, especially in
ments: customers shifting to greater reliance on debit cards
the near-term, is complex and less than completely sanguine.
may still wish to have the option of paper checks, for example. Under these circumstances, banks need to ascertain the most
Payments-related revenue is an important source of income
profitable balance between introducing new payment options
for the banking industry, perhaps more significant than gener-
and maintaining (and, over time, possibly de-emphasizing)
ally appreciated, and payments undergird both deposit and
older ones. This balancing act requires making investments in
lending relationships for banks. A large part of payments-
new payment systems, but also continuing to incur costs for maintaining older ones.19 A corollary is that banks face
related revenue comes from check-related activities, and as check use declines so will the revenue from it. It is not clear
increased risk-management responsibilities in such areas as
that revenue from increased processing of lower margin elec-
security (including fraud detection and prevention) related to
tronic payments will fill the gap. In addition, changing trends
all access points and payments options.
in retail payments have resulted in significant changes in payment fee structures for banks, including in particular a recent decline in some card-related interchange fees. Finally on the revenue side, non-banks, rather than banks, dominate markets for some higher margin, non-traditional payments options, such as remittances and bill payment, which have received a boost from recent technological innovations. On the cost side the story is also complex. While electronic payments are far less expensive to process than paper checks, some cost-related factors cut the other way. For example, as check use falls, scale economies from processing large volumes of checks decline. Even technological innovations in ‘electronifying’ checks entail expenses that must be
102
16 NACHA (an industry group of electronic payment system participants) is working to develop a mechanism to address this disparity by having banks that originate unauthorized ACH debits to accounts pay a return entry fee that will go to the receiving bank. 17 The Federal Reserve’s Regulation E covers consumer compliance issues for many types of electronic payments. 18 In fact, the experience of these banks has been that many Internet banking customers conduct more transactions, across more channels than non-Internet customers, and hence are costly to serve, despite their use of lower cost delivery
channels. However, net revenue for customers who use Internet banking tends to be higher, because these customers use more high value-added products and services, and because they tend to be more ‘sticky’ – that is, they do not tend to leave the bank, in part because of costs (in time and effort) they incur in setting up services such as electronic bill payment, account aggregation, etc. 19 For example, a May 2004 report by Celent estimates that internal information technology spending on non-card payment systems will grow 37 percent over 2003-2004, more than twice the increase over 2001-2002, and that massive investment in check processing is driving this short-term spike.
Technological innovation in retail payments: Key developments and implications for banks
Banks’ payments-related costs
In addition to the obvious start-up cost considerations, there
The common perception is that changing retail payments pat-
are two possibly less well-recognized costs of check electroni-
terns have been an important positive development on the
fication with which banks must deal. Payment security issues
cost side for banks. There is certainly justification for this
pose important cost challenges for banks. For example, many
view. Electronic payments in general are substantially less
check fraud protections are currently based on, some even
costly to process than paper-based ones. For example,
imbedded in, paper checks. Check electronification innova-
Federal Reserve per-item payments processing costs, which
tions need to be designed to offer at least a similar level of
are illustrative of industry trends, show that it is almost five
fraud protection, a process that may entail additional costs.
times more costly to process a paper check than an ACH
Additionally, it is possible for banks to underestimate payment
transaction. Furthermore, as the volume of electronic pay-
system upgrade costs because they fail to realize the need to
ments processed increases, economies of scale and addi-
properly integrate paper check and electronic (ACH) process-
tional innovations are likely to enhance this advantage. In this
es. In particular, because check and ACH processing systems
vein, Figure 7 shows that Federal Reserve System’s per-unit
have not traditionally been linked, it is possible that, when
ACH processing costs have declined as the volume of ACH
checks are converted to ACH, cash management risk controls
payments it handles has increased.
in place for paper check processing may be inadvertently
However, as with the revenue side of the picture, the story is
not be acted upon. Banks’ check electronification cost calcula-
more complex than commonly appreciated. As a consequence,
tions need to take into consideration the proper integration of
achieving overall payments-related cost-reductions entails
check and ACH processing systems.
bypassed, or stop-payment orders for converted checks may
significant challenges for banks. One fundamental factor operating to constrain payments-processing cost improvements is
An additional factor on the cost side that will tend to mitigate
declining scale economies. In particular, as check use declines
retail payments-related cost improvements for some banks
and the scale of check processing for many banks decreases,
arises from imbalances in the distribution of costs associated
the per-unit cost of check processing rises. As Figure 8 shows
with some forms of electronic payments. Such imbalances
for example, the Fed has seen a steady increase in its per-unit
may become more onerous as a growing volume of retail pay-
check processing costs as the volume of checks it processes
ments become electronic. For example, in the case of unau-
has declined.
thorized ACH payments, the costs are borne by the receiving bank (the consumer’s bank), even though the revenue associ-
paper check processing, industry participants have developed
ated with the payment goes entirely to the originating bank (the merchant’s bank).16 Other ACH-related cost burdens
technological improvements. ‘Check electronification’ results
receiving banks may incur without sufficient compensating
in substantial cost savings, including of course the elimination
ACH revenue include compliance costs, for example in connec-
In response to the increasingly unfavorable economics of
of the expense and time of moving paper through the system,
tion with consumer protection, as well as costs associated
but also efficiencies such as lower back office costs for researching past payments.15 Nevertheless, there are sub-
with screening cross-border ACH transactions in compliance
stantial cost considerations to take into account even in the
with OFAC (the Treasury Department’s Office of Foreign Asset Control) requirements.17
case of such technological advancements. These of course include equipment, software, and personnel-training start-up
Finally, there is an important strategic issue banks must con-
costs for switching to new processes.
sider that has both cost and revenue dimensions. Banks have learned that when new service options for customers are
103
Future
Innovation on networks: Coordination, governance, and the case of VISA The evolution of currency Automating payment processes to reduce working capital New agenda for payment service providers New models of collaboration in transaction banking European payment service providers: A race to market and the nature of future victory? Network-based payments and e-settlement Banks’ strategies for payment services: Which role for debit cards? Delivering migrant workers’ remittances Check 21 and the migration to electronic payments
Innovation on networks: Coordination, governance, and the case of VISA Matthew Cardillo Assistant Economist, Federal Reserve Bank of Kansas City Antoine Martin Economist, Federal Reserve Bank of Kansas City Michael J. Orlando1 Senior Economist, Federal Reserve Bank of Kansas City up problem limits innovation in a particular network setting is
ment of innovative interchange technologies. This coordina-
critical to identifying ways to enhance network utilization.
tion of investment — in this case the consolidation of research
Although the network effect has been more widely considered
and development efforts — gave rise to a number of key inno-
in these environments, solving the hold-up problem may be
vations, from BASE 1, the world's first global electronic card
the critical path to increased network utilization.
authorization system, to the adoption of the name and brand VISA and a common front-end for its products.
The case of VISA This section describes the evolution of the VISA network to
With the adoption of the VISA brand, the need to issue infor-
illustrate how network utilization was increased in the pres-
mational material and reissue all cards in circulation repre-
ence of both network and hold-up problems. We argue that
sented substantial costs for the participants to the network.
the adoption of a joint ownership structure in the early 1970s
Such a costly innovation could only have been justified if all
was critical to VISA’s growth and success.
participants coordinated to invest in the brand. These invest-
The VISA card, as we know it today, began its life in 1958 as
network. There is no obvious way for an institution deciding to
the BankAmericard, Bank of America's revolving-credit prod-
leave the network to redeploy the product of this kind of
uct and one of the first modern multi-purpose credit cards. In
investment to an alternative use. Hence, this example is also
the beginning, the BankAmericard program was entirely
typical of the kind of innovation that is at risk of being held up.
ments also have the characteristic that they are specific to the
under the control of Bank of America. In order to expand the
106
program, Bank of America formed the BankAmericard Service
The effort to consolidate research and development and coor-
Corporation, which allowed licensed banks outside the state of
dinate investments can be seen as an attempt to address the
California to issue BankAmericards to consumers in their
network effect. In this network structure, important innova-
regions. This can be seen as a first step towards mutual own-
tions should be adopted by all participants jointly to guaran-
ership of network assets, a progression that culminated in
tee the best possible return for a given amount of investment.
1970 with Bank of America's decision to transfer ownership of
However, it could be seen as puzzling that BankAmericard
the BankAmericard program, and the underlying network, to
Service Corporation would take the additional step of sharing
the banks that had been licensed issuers under the previous
ownership of the network, since it does not obviously serve
arrangement, creating National BankAmericard Inc. (NBI).
the purpose of better coordination of investment. Given the
Thus, the association, the innovative organizational structure
value of the VISA brand today, one could argue that it was a
that has played such a large role in the growth of the network,
major mistake to surrender sole ownership of such a valuable
was born.
asset. We argue that this was no mistake, or coincidence.
Apart from joint ownership of network assets, a significant
If one takes the hold-up problem seriously, it becomes appar-
feature of this organizational structure was that it called for
ent that VISA could not be worth what it is today under a
the centralization of functions that benefited all network
monopoly ownership structure. This is because the numerous
members jointly. In fact, one of the most important centralized
costly investments necessary to enhance the brand not only
functions of the newly-formed organization was the develop-
require the participation of all network participants but also
1 We thank Jamie McAndrews for helpful comments. The views expressed here are those of the authors and not necessarily those of the Federal Reserve Bank of Kansas City or the Federal Reserve System.
The ‘network effect’ is perhaps the most thoroughly studied
exceeds its benefits. Consequently, a network effect is an
feature of network environments. Participants are unlikely to
important barrier to both utilization and innovation on net-
use a network unless others are already doing so. In many
works. Unless it is possible to coordinate the behavior of par-
network settings, such as telecommunications, payments, and
ticipants, many potential network innovations will never see
software, innovation also appears fundamental to growth. As
the light of day.
such, the network effect also structures incentives to innovation that determine the network dynamic. The higher the level
The hold up problem presents a more general barrier to inno-
of innovation on a network, the more participants are attract-
vation. It plays a major role in limiting large investments spe-
ed to it. And the more participants there are on a network, the
cific to a particular application. These features are character-
greater potential for innovation. But another barrier to invest-
istic of investment in innovation on sponsored networks.
ment is likely to play a role in network growth. Investment
When an innovation is costly and specific to the network —
‘hold-up’ generally describes the disincentive to incur a fixed
meaning that there are no uses for this innovation outside of
cost where the value of such an investment may be appropri-
the network — innovation can be held up by the owner of the
ated by another party. This may be a problem for investments
network. Consider such an innovation that is adopted by all
in innovation on owned or ‘sponsored’ networks, where a set
participants to a network. The benefits from the innovation, in
of standards or rules are established by the network owner.
the form of increased network usage and/or lower cost, for
Indeed, on such networks investments in innovation are typi-
example, yield increased profits to network participants.
cally network-specific. Consequently, it is interesting to con-
However, once the cost of the innovation has been sunk, the
sider these problems in abstract and ask how one large net-
owner of the network can appropriate these benefits, for
work was able to overcome them.
example by repricing network access. Potential adopters will
A network effect results when utilization of a network service
innovators will hold up investment as they anticipate no par-
by one consumer affects the utilization choice of another. The
ticipants will adopt.
be deterred as they anticipate the owner’s behavior. Potential
telephone is the classic example of this phenomena; the benefit to using this technology depends on the number of other
In order to overcome the hold-up problem, the network owner
consumers who already do so. This problem for network utili-
must be able to credibly commit not to expropriate the value
zation translates into a similar problem for network innova-
of the investment from participants. In principle, this could be
tion. Innovative investment may not be justified if too few
done in a variety of ways. For example, a reputation or the use
ultimately use the network, but may be justified if many do so.
of long-term contracts might allow this kind commitment.
Imagine a situation where, if all network participants choose
Such mechanisms may take a long time to establish or require
to adopt an innovation, the innovation-enhanced network
costly revisions as condition change. A particularly effective
attracts even more users. In this case, benefits to participants
way to overcome the hold-up problem is for network partici-
associated with the increase in network usage may justify
pants to jointly own the network. Such a governance structure
adoption and hence investment in the innovation. However, if
insures the same set of institutions receives the benefits from
too few participants adopt the innovation, network usage may
the innovation, either as participants in or as owners of the
remain sufficiently low so that the cost of the innovation
network. Understanding how the network effect and the hold-
107
are specific to the network. Hence, in order to extract more
tional structure of VISA and American Express. Either all par-
value out of its network it was necessary for Bank of America
ticipants jointly own the network, as a way to solve the hold-
to agree to share it with the other banks that licensed the
up problem, or a single entity is in charge of a ‘network’,
credit cards.
integrating the interests of downstream participants into the firm.
Imagine for a moment that, in the early 1970s, BankAmericard Service Corporation had proposed to centralize the research
As we have pointed out, the hold-up problem is likely to be
and development function for the VISA network but retained
particularly serious in cases where costly, network-specific
sole ownership of its assets. In a world where the only barrier
investment is necessary to implement innovations on a spon-
to innovation is the network effect, this step would have been
sored network. Such investment cannot provide benefits to
enough to spur the growth of VISA. However, we have argued
participants outside the network and thus makes them espe-
that one should expect other participants in the VISA network
cially vulnerable to being expropriated by the network owner.
to realize that such an arrangement would be bad for them. In effect, they would be financing costly innovations that enhance the value of the network but would be unlikely to partake in those benefits. BankAmericard Service Corporation, as the sole owner of the network would be able to reprice network access to systematically extract all benefits. This suggests that the only way to get other card issuers to agree to finance network-specific innovations was to allow them to become joint owners of the network. One more point is worth noting. At first glace, the governance structure of American Express might seem to provide a counter-example to our story, since, in contrast to VISA, American Express is not mutually owned. Closer inspection, however, reveals that there is no contradiction. American Express, because it is a proprietary network, and not an association, has no independent banks issuing American Express credit cards. Instead American Express functions as its own sole issuer, meaning that, unlike VISA, it is characterized by a vertically integrated organizational structure.
Conclusion In principle, by the logic of the network effect argument, coordination could have solved BankAmericard’s network problem. In practice, BankAmericard had to take an extra step and organize under a mutual-ownership governance structure in order to address the hold-up problem. Hence, our argument suggests that there is no middle ground between the organiza-
108 - The
Journal of financial transformation
109
The evolution of currency James Turk Founder, GoldMoney.com
Thus, the concept of a money substitute introduces a new risk
tutes they create to circulate as currency. That is a high cost
into the transaction, normally called payment risk. It is the risk
indeed to consumers and businesses, particularly when con-
that the recipient may not be able to exchange the money
sidering that the technology exists today to significantly lower
substitute received in a transaction for a good or service. The
these costs.
payment risk means that the money substitute may lose all or some of its value before it can be exchanged for items of
I expect, therefore, that currency will evolve into new forms.
value.
And, it will not be the first time that technology has caused
For more than 300 years we have been using money substi-
1977 corporate IT departments ‘ruled the roost’ and PCs were
profound changes. Take the history of PCs as an example. In tutes. Their problems are readily apparent. Paper currencies
a hobby for a handful of enthusiasts. Yet in ten years, the PC
often become worthless if banks fail, or when central banks
had turned corporate IT departments on their head. Because
pursue reckless policies that erode, and in some cases destroy,
of developing hardware and software technology, the PC had
the value of the currency.
become an important new tool, forever changing the face of corporate computing. This same potential exists today for the
Furthermore, money substitutes are expensive. They are
Internet to fundamentally and forever change bank payment
based upon credit, and it is costly to monitor credit-worthi-
systems. As a consequence, look for new companies to exploit
ness. In addition, substitutes do not circulate efficiently in
these new technologies, and in the process create a currency
online commerce. There are costs when converting one cur-
that once again is money, and not a money substitute.
rency to another, and more importantly, with national currencies it is not possible to make instantaneous and nonrepudiable 24/7 payments, which have clearly become important requirements for global online commerce. In short, while money substitutes have more or less met the needs of the marketplace up to now, they have become inadequate. Importantly, modern computer and communications technologies make possible new forms of currency, even private freemarket currency, such as GoldMoney. The digital gold currency is driven by bottom-line objectives to earn a profit, in contrast to the often capricious and harmful political aims under which central bankers operate. Therefore, it is my expectation that currency is about to evolve yet again. Global online commerce requires it, and so does the high cost of operating our present currency system of money substitutes. Banks in the U.S. earn approximately U.S.$70 billion annually, and it is estimated that some 40% of their profit is derived from their payment systems that enable the money substi-
110 - The
Journal of financial transformation
Until the formation of the Bank of England in 1694, money and
holders of the notes some certainty that their money would be
currency were one and the same. Money was a tangible asset
returned to them, but it also introduced a risk, namely that the
and this asset itself circulated as currency.
bank would default on its obligation.
Money most often took the form of gold or silver coins, though
This risk of default was generally perceived to be acceptable
history shows that many assets have from time to time served
given the advantages offered by this new currency. Because
as money. But regardless of what ‘thing’ served as money,
bank notes were very convenient and lowered the cost of
each shared one characteristic; they were all a tangible asset.
transactions, they offered significant advantages over coins.
Consequently, money had value in transactions because the
Therefore, increasing amounts of bank notes were put into
asset being used to represent it was perceived to possess
circulation. It was a successful innovation. However, the essen-
some inherent usefulness and was therefore valuable. What is
tial nature of currency was radically altered.
more is that it was this tangible asset that circulated as currency. In other words, though money often took different
In contrast to the tangible asset commodities that circulated
forms, with coins of precious metals being the most common,
as currency, bank notes do not have any intrinsic value. Their
the asset serving as money passed current from hand-to-
value is derived from the assets backing them, thus making it
hand.
clear that currency had become an item that requires balance sheet accounting and double-entry bookkeeping. Currency
Assets serving as money, however, were inconvenient. Coins
was no longer a tangible asset, the value of which arose from
were heavy and bulky, and the large number required to com-
the asset itself. Currency was now a money substitute, and in
plete high value transactions was unwieldy. Even worse, coins
particular, a liability of the Bank of England, which promised to
could easily be nicked or scratched in use and were worn down
pay money on demand.
from constant handling, dissipating the precious metal content that gave them their value. As a consequence, gold and
There is an important difference between money and money
silver were considered too valuable to circulate as currency, a
substitutes, which has its basis in common law, that for an
reality that acted an as incentive to develop alternatives. And
exchange in the marketplace to be extinguished assets have
in fact these limitations were overcome by an important tech-
to be exchanged for assets. So if one uses gold to buy a horse,
nological innovation from the Bank of England.
for example, an asset (gold) is being exchanged for an asset (the horse), and the instant the assets change hands the
It created paper currency, called banknotes, which were a
exchange is extinguished. There are no lingering obligations.
promissory note obligating the bank to pay the weight of
But consider this transaction if one uses a national currency,
metal printed on the note. This new currency enabled the gold
i.e., a money substitute.
and silver to remain in the bank’s vault, with these paper notes circulating as currency in place of the inconvenient and valu-
The national currency is not a tangible asset. It is a deposit
able metals.
liability of a bank, and it derives its value from the assets backing it. Therefore, the buyer in this exchange walks off with a
This new currency was therefore a money substitute, and not
horse (an asset) and the seller walks off with a money substi-
money itself. The currency allowed the appropriate weight of
tute (a bank’s liability), so the exchange is not extinguished.
precious metal to be paid on demand when the bank note, on
There is a lingering obligation, and the exchange is not extin-
which a given stated weight of the metal was represented, was
guished until the seller manages to exchange the national-
handed to the bank for redemption. This feature provided
currency money-substitute for a good or service.
111
Automating payment processes to reduce working capital Kurt Cavano Chairman and CEO, TradeCard
Executives there were looking to automate the payment deci-
required to precisely manage and exploit this information.
sioning and scheduling processes. The retailer chose not to
Such tools can provide everything from a big picture view of
implement software across its procurement division and hun-
payments to the status of individual transactions. Having this
dreds of suppliers, but instead to use an on-demand platform
information allows finance teams to lower the working capital
that works over the web. The platform extracts purchase
they keep on hand and, in turn, apply funds to higher yield
orders from the retailer’s ERP system, and then automatically
investments. It also improves planning for currency exchange.
populates and routes other documents necessary for the
The result, in short, is the ability to fund the right payments at
transaction, such as suppliers’ invoices. It also handles all the
the right time with the right currency.
information gathering and compliance checking that goes into payment decisioning and scheduling, giving the finance team the freedom to think strategically about payments and their timing. From a technical standpoint, what makes this possible is open data exchange. A solution, whether it is software or an on-demand product, should be able to accept nearly any data format and transport, from EDI and AS2 to XML and flat files. That way, all the work a company and its suppliers have done on procurement, fulfillment, payment, and workflow is reused. Likewise, both buyers and suppliers can continue working with their existing logistics and financing partners.
Use procurement and payables data to reduce working capital Poor visibility into and control over payments shrinks the capital a company has available for other uses. With the right solution, companies can track relevant data throughout the procurement cycle, enabling finance teams to precisely forecast cash movements, plan for FX, and better invest the company’s money. For instance, the finance team of a company I know uses their procure-to-pay automation platform to leverage advance shipping notices. Suppliers issue these notices when the purchased goods are ready to ship from their factory or warehouse. Based on the notices, the finance team begins forecasting payables much earlier than they could with invoices. Other companies move even closer to the purchase order. Many continue to rely solely on invoices, with the added benefit of an automated mechanism for collecting the data. In any case, a dashboard or some other reporting tool is
112 - The
Journal of financial transformation
When CFOs and other finance executives discuss electronic
Most companies begin automation at the purchase order, and
payments, the conversation typically centers on optimizing
look to electronically share this information, as well as logistics
settlement. But what about the preceding steps, particularly
and payment information, across their own accounting and
payment decisioning and scheduling? Payment decisioning
procurement departments and with suppliers. The advantages
determines whether a purchase order’s terms and conditions
of starting automation at the purchase order are twofold.
are fulfilled and reflected on the invoice — and in turn, wheth-
Firstly, this approach contributes to overall process efficien-
er the invoice should be paid or adjusted. Payment scheduling
cies, as information is disseminated faster and more accu-
encompasses the payment’s timing. Both areas require pre-
rately. Secondly, finance teams gain the option to take pay-
cise information, yet are typically rife with manual processes
ment visibility to its very genesis. With the right solution, pay-
and paper. The result is reduced visibility into payables and
m
bloated working capital levels. These issues are exacerbated
visibility can even be aimed at logistics areas, such as when
e
n
t
when procurement is decentralized, such as at a company
suppliers ship goods or when a warehouse receives them.
whose multiple divisions use individual purchasing and accounts payable processes.
There are several software and on-demand products that automate these ‘procure-to-pay’ processes. No matter which
In contrast, when payment decisioning and scheduling are
one an organization chooses, it is important that the solution
executed efficiently, companies can tightly control which sup-
be able to base payment decisioning on the comparison of
pliers get paid on what dates, as well as extend visibility into
multiple documents and sets of information. After all, many
upcoming payment requirements. This complements elec-
factors can be involved in determining whether or not a sup-
tronic payment’s process savings while enabling companies to
plier has met their obligation.
reduce working capital. For these reasons, an analysis of payment decisioning and scheduling processes should be part of
Another key feature is dispute resolution and negotiation.
electronic payment initiatives.
Having automated workflow and collaboration capabilities in these areas can expedite the handling of non-compliant
Many companies determine that an automated approach to
invoices or inaccuracies in other documents, saving time in
payment decisioning and scheduling is a sensible solution —
accounts payable as well as procurement. It is a particularly
particularly if the automation layers over existing business
important capability for direct goods procurement, as slow-
processes and technologies, rather than requiring changes to
downs anywhere in the supply chain can impact production
them. Below are three best practices for implementing this
schedules and the cost of goods.
approach. They are based on the practical experience of finance executives who led similar projects.
Extend the value of ERP and other back-end systems
Start with the purchase order
Many companies have put a tremendous amount of resources
Payment decisioning involves comparing procurement, fulfill-
into ERP and other back-end systems. To expedite procure-to-
ment, and even warehouse documents to determine if a sup-
pay automation, it makes sense to leverage these invest-
plier’s invoice should be paid. Automating this tedious process
ments. A good solution will reuse their data, such as purchase
gives finance teams faster notice of upcoming payments. By
orders or shipping notices, and complement their associated
adding automation of payment scheduling, companies can
business processes.
then settle invoices at the most financially opportune moment. Consider the case of a large retailer I recently visited.
113
New agenda for payment service providers Terry Dirienzo, Director, Group Payments & Settlements, Barclays Plc. Mark Hale, Head of Payments Strategy, Group Payments & Settlements, Barclays Plc.
support these developments. As customers invest in more
that optimize their supply chain by combining a competitive
electronically integrated business processes, banks must
unit cost with excellent service delivery.
ensure that their payment services can be easily integrated.
Organizational philosophy Critically, as a customer’s expectation of service increases and
In assessing the options with which to meet a bank’s payments
their willingness to pay a premium decreases, banks must
needs, it is important to remember that payments is only one
seek to improve the productivity of their payments process-
element of a bank’s business and that any organizational deci-
ing, clearing, and settlement. This creates demands and chal-
sions for payments must take account of the wider context in
lenges for every part of the internal and external supply chain
which they occur. Central to this is whether a bank has a cen-
from product, through channel and operations, to clearing and
tralized or a decentralized structure for measuring the suc-
settlement. The ability to process payments electronically
cess of its products and services, which includes the way they
from originator to beneficiary, straight through and without
are managed and delivered. If the structure is centralized, it
expensive intervention, will be essential to driving up produc-
provides for decisions to leverage shared infrastructure, to
tivity and quality.
offer a manufacturing service for payments, and to obtain high rates of asset utilization. If the structure is decentralized,
In view of the extent of information technology employed in
then a more distributed delivery capability is possible; trading
the supply chain infrastructure, managers must take account
higher fixed cost overheads for greater flexibility and control.
of the IT processing and cost benchmarks of similar IT enter-
The ability to adopt a more externalized sourcing model will
prises, and achieve at least the same or similar levels of asset
also depend to a very large degree on the extent to which the
utilization and rates of return on capital employed.
existing payments infrastructure and operations are separated, colocated, rationalized, or optimized.
Every bank has a different payments service and product mix, as well as a different marketing strategy, which will largely
A bank’s fundamental perception of payments will dictate
determine the process and operational requirements that
how they approach this choice. In particular, for example,
need to be satisfied. Equally, banks will be at different stages
whether payments are regarded as a source of value, an
of organizational evolution, which will determine the extent
application of cost, or an element of risk. This decision is not
and nature to which an integrated end-to-end payments solu-
exclusive or one-dimensional, since the perception may differ
tion is available to customer-facing business areas. The com-
depending on which part of the supply chain is being con
bination of these factors will then help identify the most
sidered (payments, clearing, or settlement), the customer
appropriate organizational and delivery decisions for actual
segment being targeted (retail, corporate, or financial institu-
product and service delivery; in particular, identifying the ser-
tion), or the driver of the activity (interoperability or differen-
vices that should be processed internally and those that
tiation).
should be purchased externally. If payments are regarded as a cost, a non-differentiated com-
114 - The
The banks that will succeed and gain competitive advantage
modity, and a homogeneous product, then this will lend itself
under these conditions will be those that can develop out-
to a centralized structure and to more external sourcing.
standing propositions to create genuine customer value and
Whereas a more profit-oriented treatment of payments, that
Journal of financial transformation
The business of making payments should be high on the
The application of information communication technology to
executive agenda of every financial service provider over the
enable real-time transfer of value, to support daily transac-
next five years. However, it is a business in every sense of the
tional values that sometimes dwarf the GDP of many coun-
word and this means that managers have to reconcile provid-
tries, and to provide international transaction processing
ing customers with the most relevant and value adding pay-
means that the payments business has an extremely high
ment services, while at the same time doing so in the most
inherent level of operational risk. Banks, therefore, have to
cost-effective and responsible manner that mitigates the
fully understand, manage, and be able to mitigate the man-
inherent investment and business risks.
agement risks involved in their payments services.
Customers expect their banks to make payments, to have suf-
Independent legacy systems, often with proprietary technolo-
ficient payments reach, and to be able to make their cash
gies and systems, are being challenged by the consolidation
management more efficient. The business of payments is,
and globalization that is occurring in upstream clearing and
therefore, a core competence of a bank in the same way as
settlement and in downstream customer treasury and cash
lending money and providing saving vehicles are core compe-
management. This is being accelerated by standards conver-
tencies, except that it is the lynchpin of the three, since the
gence and the use of more open standards, thereby allowing
other two depend on payments to function.
greater access to new entrants and opportunities for keener competition.
The unprecedented change in the business environment, relentless increases in the level as well as the nature of com-
The economic importance of payments also means that mar-
petition in the market place, and more sophisticated customer
ket makers and regulators are seeking to achieve their goals
demands, therefore, create a demanding and urgent agenda
and ambitions by homogenizing payments activity within
for all payment service providers.
their respective countries and also cross-border. This is particularly true within Europe, where with the Single European
Payments landscape
Payments Area is being used as a means to achieve a single
Payments are also the lifeblood of commercial activity and as
European market.
such they underpin the operation and the confidence that makes every economy function. In an information technology
The challenge for banks
era, therefore, it is imperative to ensure that delivery of pay-
The payments context and landscape means that banks need
ments is efficient, safe, and trusted; irrespective of how the
to deliver payment solutions that their customers value, in
payments are actually provided within the supply chain.
the manner most attractive to those customers and in the timescale that fits their own plans and activities. As custom-
Customers have an extremely high expectation of payment
ers themselves adopt new technologies, so too must banks
service performance and a very low tolerance threshold for
adapt their payment solutions. This is particularly true of the
payment errors. This places a premium on the various ways
Internet, where corporate customers are using it to improve
that the channel, product, and service strategies of banks
their own customer experience, to broaden their proposi-
address the matter of payments and in particular in how these
tions, and to improve their workflow and cash management
products and services are sourced.
cycles. It is essential that banks anticipate and are ready to
115
how the service is delivered the risk and the responsibility will
Final comment
always remain within a bank. The combination of these three
A number of commentators talk about banks simply outsourc-
regulatory pillars may cause many banking executives to be
ing their payments businesses to reduce cost, but this seems
nervous of allowing third parties to manage substantial por-
to underestimate the operational risks involved and the value
tions of operational risks like those inherent in payments.
of payments to the business of banking. They also overesti-
Market development conclusions
commercial requirements involved. It seems more likely in the
The management of the payments supply-side is complex,
long-term that banks will, either unilaterally or in partnership
risky, and subject of substantial future change. Any third party
with others, establish a transactional processing utility to pro-
mate the ability of the supply-side to meet the full range of
seeking to provide compliant solutions will want to ensure
vide payment-processing services to other banks. In the short-
that their pricing includes a premium for these risks and will
term, the focus is more likely to be a mixed sourcing model
probably shy away from absolute warranties and assurances
with a focus on raising processing productivity and improving
about obtaining compliance. It is not the change that is the
service quality.
primary issue here, but the fact that change is certain, will be mandatory, and yet difficult to quantify. It is clear, however, that banks will continue to develop their outsourcing and offshoring experience. Given the discussions above, the short term upstream supply-side is likely to develop in the following ways. Banks will focus on making improvements to internal processing capabilities to allow more thirdparty sourcing solutions to be considered. There will be in足creased standardization of payments activities to facilitate migration to different sourcing models. More co-sourcing of common processing activities in the non-competitive areas of clearing and settlement will be undertaken. There will be greater consolidation of clearing and settlement activities, which will be accelerated by the development of the Single European Payments Area. Selective third-party sourcing of discrete and niche payments business activities will become more prevalent; this is an area ripe for joint venturing. And, third-party sourcing of infrastructure management and some operations management will continue. Knowledge management as a separate and managed discipline does not have an established history in banking and this, as well as other critical pre-requisite disciplines, will need to be established before banks are able to safely migrate substantial parts of their payments businesses to third parties.
116 - The
Journal of financial transformation
perhaps services niche and profitable payments markets, will
sourcing from competitors without non-compete and secrecy
lend itself to internal manufacturing, selling of transactional
agreements is particularly fraught and carries the risk of
services, and some decentralization of niche activities.
brand and reputation damage. Where, for example, is the com-
Sourcing models
service if they are able to approach the customer in question
Managers constantly address the question of how to source
and subsequently solicit their business?
mercial incentive for a competitor supplier to offer a good
each component of the supply chain. Depending on the scale of the component in question, this ranges from in-house deliv-
Key sourcing considerations
ery to facilities management, joint venture, or full outsourc-
Commercial considerations also make outsourcing of the pay-
ing. Within each category there are then other key consider-
ments business particularly complex, and Basle II exacerbates
ations, such as packaged or bespoke services, offsite/near-
this. There are many substantial credit exposures in the pay-
shore sourcing, or finally full offshore sourcing.
ments systems at any time during the day and these are subject to collateralization and reserve. Normal business practice
The considerations within each of these options are fairly
is to match risk with control. But, where control of payments
standard and best practices also fairly well established and
is in the hands of a third party it is hard to conceive that these
understood. Much of this is codified and available in reference
exposures could easily be covered by liability agreements, it is
texts. The issue is principally one of information systems,
easy to see the insurance pool being consumed by a market of
infrastructure, and information technology services manage-
several ‘pure’ payment service providers, and there are only a
ment. However, where there is more uncertainty, and much
few organizations with a balance sheet capable of sustaining
less guidance, is in how to apply these options to the business
the contingent liability.
of payments and in such a way that satisfies the prudential requirements of a regulated credit institution. The inherent
Additionally, given the mission critical nature of payments
and systemic risks involved in the payments business is of
processing, the level of inherent risk, and the fixed operating
such an order of magnitude different to other business
deadlines for clearing and settlement where services are out-
activities in other industries that the Committee of European
sourced, there is a premium for excellent service integration
Banking Supervisors has established principles for outsourc-
and issue management capabilities. This is particularly the
ing material activities, made such decisions the subject of
case in a mixed sourcing environment. In addition, where a
pre-notification, and prescribed the outsourcing of certain
bank is accustomed to understanding the imperative of sys-
specified activities.
temic and regulated stability in issue management and used to placing these considerations before immediate commercial
At a more component level, banks already utilize a mixture of
concerns, there could be a risk when there is substantial third-
different sourcing solutions to meet their needs. This may be
party sourcing.
to take advantage of scale economies or to white label the delivery of particular software products. In every case it is
Under Sarbanes-Oxley, bank executives are more clearly and
essential that the customer is unaware of this and that the
directly accountable for the effectiveness and operation of
actual supplier of the product or service is not transparent to
their internal control environments. Under Basle II more
them.
effective operational risk and incident management will be
Payments is a relatively homogeneous product in isolation
capability. The Committee of European Banking Supervisors
and therefore subject to supplier switching. This means that
does not allow risk management to be outsourced. No matter
required, as well as a much better information reporting
117
New models of collaboration in transaction banking Wolfgang Gaertner Chief Information Officer, Global Banking Division — Global Transaction Banking, Deutsche Bank, and, Director, SWIFT dollar clearing market has seen the disappearance of 59 of
also frees up banks to move further up the value chain to
the 115 clearing house participants during the last ten years
meet the developing needs of their corporate clients. Banks
— more than 50 % of the market players have disappeared
can look at how and where they can continue to add value,
(CHIPS statistics). This supports the hypothesis that as the
especially for multinational corporations, by helping compa-
consolidation process of the transaction business continues,
nies to manage liquidity across their global operations. Thus,
there will only be a few banks that will have the transaction
these new business models, with cooperation between all
volume, and hence the scale of economy at the lower transac-
types of institutions, could allow for more efficiencies and
tion revenue margins, to meet the customer demands. The
lower costs for all.
rest need to think about new ways of doing business — coming back to the car builder analogy, thinking about breeding cattle
One example for white-labeling is the cooperation between
or outsourcing the farm.
Barclays Bank and Deutsche Bank. Such strategic partnerships could only be possible due to excellent IT and operations
New ways of partnering between banks
capabilities, and shows that these fields have become key dif-
Outsourcing to another bank may not only serve to reduce
ferentiators and critical success indicators. The mandate from
costs, but can enhance the existing offering to customers
Barclays came with the responsibility to provide premier ser-
wanting more complex services. It may even improve the cost
vice, best quality and highest system availability to the bank
basis of certain parts of the back-end value chain. There are
and its clients. This is the first partnership of its kind in
possibly three wholesale banking products and services that
Europe, where a leading provider offers its cash management
are part of this trend:
franchise to another financial institution to serve its corporate clients. By leveraging Deutsche Bank’s infrastructure and
■ White labeling of front-end technology — namely Internet-
based services, such as electronic banking platforms and netting systems. ■ Infrastructure outsourcing of IT and operations with trans-
technology investments, Barclays Bank can now offer state-ofthe-art pan-European cash management solutions to clients, including account services, domestic and international payments and collections, liquidity management, and electronic
action routing and conversion capabilities, transaction
banking solutions. They will, of course, continue to provide
settlement options and lockbox services being popular
account and payment services to their customers within the
choices.
U.K. and retain full responsibility for client relationships, while
■ Partner banking solutions, including account and liquidity
Deutsche Bank will oversee operations and processing, the
management services, infrastructure and platforms
management of its services, and also extends support to
around common services, such as sales, implementation,
Barclays’ sales, implementation, and customer service teams.
and customer service models. In such a scenario, there are winners all around the table.
118 - The
Fundamental to partnerships is the selection of a suitable sup-
Banks unable to maintain a stand-alone cash management
plier. Key benchmarks are a commitment to quality, price,
business can buy-in services to protect their customer base,
reputation, and resources. Clients would see little difference,
reduce their cost and risk, as well as generate revenue. The
as such banks would provide the same depth of the value
vendor bank leverages its infrastructure and systems invest-
chain, by outsourcing services to larger institutions. Thus,
ments, as well as its high-volume processing efficiencies, to
banks could maintain their relationships and client interaction
create an alternate stream of revenue. Finally, corporations
without the cost and effort of managing the entire process. It
and institutions stand to benefit from access to products and
Journal of financial transformation
In the world of transaction banking, change has always been
Now that transaction banking has become a commoditized
incremental and not fundamental. But not anymore, the
business, with transaction fees decreasing, clients look for
industry is transforming itself, and at great speed. It is hard
lower prices and increased capabilities. Yesterday’s order-
to believe, but instituting such fundamental changes will
winning criteria are just normal today, so the requirement to
force some banks and financial institutions to move away
deliver truly complex solutions and the ability to process
from transaction banking altogether, which was once regard-
scale has increased significantly. In addition, a difficult global
ed to be a core business. These necessary changes, when not
economy and a harsher regulatory environment are putting
totally successfully implemented, can become a costly pro-
more pressure on banks. The ever-increasing number of regu-
cess and involve high operational risks.
lations is an additional cost driver. For example, TARGET 2 will result in the realization of a single real-time settlement sys-
Now what exactly are the drivers for change in transaction
tem for Europe. Similarly, the anti-money laundering legisla-
banking? We have increasing competition within a declining
tion and the U.S. Patriot Act will require further investments.
margins scenario, a growing set of demanding regulations,
And while large, global providers have the systems in place to
greater customer expectations, massive IT investments and
comply, smaller banks will find it hard. Both the Patriot Act
growing expertise as a prerequisite, new models of collabora-
and the Sarbanes-Oxley Act place emphasis on internal con-
tion in the financial industry sector, and the resulting industry
trols, transparency, and traceability of data. Furthermore, the
consolidation. Understanding these drivers is key for the
U.S. Check Clearing for the 21st Century Act (Check 21) means
finance director and corporate treasurer, as they will have an
that most checks now need to be converted into images and
impact on how a company manages its financial transactions.
many payments will take the form of e-checks. Market forces
It is these very drivers that will also change the transaction
and regulatory changes, such as the E.U. regulation on cross-
banking landscape dramatically.
border payments in Euroland, affect transaction processing profitability. At the same time the establishment of SEPA, the
While competition is increasing, the transaction market will
Single Euro Payments Area, results in the necessary develop-
undergo a strong consolidation as banks try to come to terms
ment of new and better systems. IT specialists, better
with all the changes that our industry is facing. They look at
straight-trough-processing (STP) rates, and investment gov-
what services they should and are actually able to offer to
ernance and control processes are needed.
their corporate and institutional clients. It will be increasingly difficult and investment-intensive for many transaction banks
As the requirement for optimization and pressure on banks to
to offer the complete suite of transaction services in-house
achieve fast operational efficiencies continues, smaller banks
and manage the investment that goes hand-in-hand with
must individually evaluate their capabilities. This could
that. The result: nothing less than paradigm shifts.
require some painful decisionmaking before proceeding with capital investment to enhance existing or purchase of new
There is the old saying about car manufacturers, that at some
payment applications to support new IT infrastructures, or
stage their vertical range of manufacturing would get so
even to enhance STP.
extensive that they would eventually breed their own cattle in order to have the leather for car seats. While it is clear that
Should banks, therefore, maintain payment processing in-
this situation has changed dramatically in manufacturing, it is
house, make a strategic decision to set up a transaction bank
less obvious that this is what banks are doing in today’s mar-
and in-source, or make a strategic decision to outsource to
kets when it comes to transaction banking.
one of the global payment providers? For example the U.S.
119
ment instructions received from our customers), which could
tions. Therefore, they have to be clear in their decisions on
not be processed straight through and require manual inter-
how to position themselves within the context of the industry
vention. With this implementation all former manual interven-
drivers.
tion functions will reside on one application, enhancing processing speed, STP, and efficiency greatly. In effect, with this
Financial institutions are requiring more and more from their
break-through installation, Deutsche Bank will run their pay-
payment service providers because their customers are
ments business on the same platform globally, and this is for
demanding more. Therefore each bank has to ask itself, what
all processing hubs in all major currencies. This single plat-
payment services can they provide and at what cost? How can
form also maximizes the impact of the desired global opera-
they make such payment processing competitive and from
tion model. The architecture behind MTNA is specifically
where do they process?
designed to support the planned key strategic initiatives, among them the full multi-entity functionality required to
The ongoing market consolidation is a result of the market
undertake in-sourcing or private label processing, artificial
necessity to offer the full global suite of transaction banking
intelligence components and learning capabilities, making the
products at competitive prices. Consequently, large transac-
repair/STP processes more efficient than ever before and
tion banking providers will be able to provide in-sourcing
providing the basis for state-of-the-art delivery of advising,
activities and white-label services for the mutual benefit of
and reconciliation and decision-support information to clients.
their strategic partners. This has resulted in consolidation among transaction bankers and has led to a greater accep-
MTNA comprises a single system with common components
tance of outsourcing as a viable option as technology, regula-
for euro, U.S. dollar, sterling, and multi-currency clearing and
tion, compliance, and contingency. There is also a growing
comes with a full array of conditional payment triggers, which
recognition that the need to meet these requirements raises
will become increasingly important in a real-time payments
the bar on the investments required for current providers to
environment. Importantly, customers will not need to adopt
remain competitive.
new technology; they interface with MTNA in a number of ways, among them SWIFT or other proprietary Deutsche Bank
Today’s transaction banker will have to recognize that the
payment offerings. MTNA is an example of how global and
demands of customers must be met. These demands cover
complete IT strategy must deliver in the days of market con-
such aspects as low transaction price, responsive service lev-
solidation and new ways of doing business in transaction
els, faster processing with the ability to accept transactions
banking, and how IT can help to differentiate a bank from its
closer to payment cut-off times, provision of accurate transac-
competition.
tion processing supported with fast research of enquiry proce-
Conclusion
covers these demands, and is an example of how IT can serve
dures, and the provision of added product functionality. MTNA When it comes to transaction banking, large technology
as a differentiator in a difficult market. It is also an example of
investments and their underlying control processes are need-
how an effective IT architecture, combined with the use of
ed just to stay competitive and to comply with regulations. For
global modules, and a strict investment cost governance over
financial institutions, the resulting consolidation leads to in-
underlying processes, have helped enhance the transaction
sourcing and outsourcing decisions at the same time.
banking capabilities of Deutsche Bank.
Especially, banks are in the dual position of being both customers of other banks as well as service providers for corpora-
120 - The
Journal of financial transformation
services they would not otherwise have been able to procure
invest in platforms and capabilities. Better back office capa-
through their house bank.
bilities and enhanced STP rates are preconditions for achiev-
Another example, this time for the outsourcing process taking
ments business. The following drivers need to be fully under-
place at the same time, is Deutsche Bank’s partnership with Postbank. On April 6th, 2004, Deutsche Bank AG and Postbank
stood when STP and back-office automation are to be
signed the agreement that the latter takes over the process-
ization trends; streamlining processes and systems across
ing market concentration and standardization within the pay-
increased to competitive levels: Assessing back office rational-
ing of the former’s domestic payments as well as the paper
regions to deliver benchmark cost transaction processing;
handling of Deutsche Bank’s cross-border payment opera-
understanding the upcoming challenges for central and com-
tions. Postbank now processes the German domestic and
mercial banks arising from the E.U. accession; restructuring
parts of the foreign payment transactions for us. For both
the back office and delivering benefits through outsourcing
banks, this cooperation is a step in the implementation of
and in-sourcing; examining the current drivers for STP, assess-
respective strategic objectives. Postbank is continuing to
ing the latest developments within E.U. payments regulatory
expand its domestic transaction banking business, leveraging
requirements; and functional and analytical clarity on effi-
their resources, while Deutsche Bank will be able to concen-
ciency gains, expense control, reduced error rates, increased
trate on their core transaction banking businesses. For the
scalability, and operational risk.
banks’ clients, nothing will change. Their accounts will continue to be kept, as before, with their respective banks.
It is a huge and complex task to provide efficient and robust solutions across different regulatory environments, real-time
Postbank is already one of the market leaders in payment
reporting, investigations and analysis requirements, tax
transactions with approximately 10 million transactions per
issues, and formats and clearing mechanisms, to name a few
day. Based on the agreement to handle Deutsche Bank’s pay-
of the challenges. For example, the Global Messaging Archi
ment transactions, and this is interesting, along with Dresdner
tecture (GMA) project addresses these requirements, while
Bank’s payments, agreed upon in March 2004, Postbank will
Money Transfer New Architecture (MTNA) is a global initiative
have a market share of over 15% of the domestic payments in
to replace segmented regional High Value Payments process-
Germany. In return, Postbank will transfer to Deutsche Bank
ing systems in Germany, the United States, and the United
its U.S. dollar clearing business and its worldwide foreign
Kingdom with global modules. MTNA is a new state-of-the art,
transaction banking outside of the E.U. Deutsche Bank is thus
component-based processing system. The major component
strengthening its core competencies in offering global cash
of the U.S. release, which is now up and running, is an
management solutions and the settlement of foreign payment
enhanced liquidity management and clearing module. This
transactions, in particular for institutional and corporate cli-
manages the gateway to the Federal Reserve Bank and pro-
ents. As a result, both parties are advancing by taking con-
vides tools to monitor and manage intra-day liquidity, clearing
crete steps to achieve joint synergies in transaction banking.
channel accounts, and participant positions. The Party Search Engine component is a look-up tool that provides speedy
Technology as an enabler
access to bank information needed to qualify and repair pay-
As mentioned before, in transaction banking, IT has become a
ments.
huge differentiator. Cutting costs by enhancing STP, reducing operational risk, and catering for global platforms have all
The manual entry and repair system module includes func-
become paramount. Major players will need to continue to
tionality to repair and approve outward payments (i.e. pay-
121
European payment service providers: A race to market and the nature of future victory?1 Julian Wakeham, Partner, Capco Andrew Hogan, Managing Principal, Capco Hector Nelson, Consultant, Capco Payments have long been central to a bank’s value proposi-
unit pricing, with higher unit costs often resulting from the
tion. However, on-going customization of the payments indus-
variety of mechanisms in place.
try, the advent of the euro, and the emergence of a Single Euro Payments Area (SEPA), are now forcing European banks
Consequently, the demand for a shared payments business
to reconsider exactly how their payments propositions should
service provider (BSP) platform, with sufficient scale to obtain
evolve. Emerging infrastructure standards, such as SwiftNet
efficient cost parameters, has emerged. These emerging mod-
and EMV, payment solutions, such as EBA Step 2 and
els are creating a market for outsourcing and driving financial
VisaDirect, and both ‘e’ and ‘m’ channels are all forcing pay-
institutions to invest in innovative, revenue generating new
ments players to face up to some key questions relating to
ventures. These BSPs offer full service capabilities, delivering
their business strategy in the years to come.
both business processes and technology services to third parties and are run by people with experience in operating the
In order to remain competitive, European banks are being
business. Such platforms would require the cooperation and
forced to make strategic decisions about where and how they
collaboration of a number of nationally dominant players,
will operate within the different ‘value webs’ currently emerg-
working in concert toward a common goal. To date, franchise
ing across the landscape. The rise of more customer-centric
conflict, substantial implementation costs, and a number of
payments systems are forcing banks to reexamine their pay-
social challenges, particularly in continental Europe, have all
ment value propositions in light of both changes to market
proved to be substantial barriers and thus far the cost of these
infrastructures and the increased competitive threat posed by
have outweighed the benefits of large-scale bank collabora-
fast-moving new entrants. The time would appear ripe for the
tion in the payments space.
creation of a payment utilities model, designed to allow banks to share costs, minimize risks, and better utilize funds that
Some banks have dipped their toes into the BSP waters, but
would have previously been sunk into payments infrastructure
genuine success stories are few and far between. In fact, a
purchases. Yet, the evidence shows that very few utility mod-
review of past BSP ventures reveal a results spectrum ranging
els thus far have managed to gain significant traction in this
from underperforming attempts like European Transaction
market. While this track record has not impressed the pay-
Bank (ETB), Finforce, and Arcordia, through to partial success
ments market to date, there are signs that the situation is
stories like Proponix. Additionally, there are a few emerging
changing and that the reality of a successful utility is starting
ventures where the verdict is still unclear. BSPs that have
to materialize.
failed to build the necessary momentum for market success
Why BSPs have underperformed so far
of client banks required for such a scale-driven business
Historically, the payments market has been characterized by
model. In fact, they have never achieved the market credibility
have essentially not succeeded in attracting the large number
high levels of competitive concentration, with a small number
needed to build a critical mass of participant transaction vol-
of large banks dominating domestic payments processing and
ume. The Proponix venture has enjoyed some momentum, but
clearing businesses. These structures have traditionally been
challenges around market pricing have prevented it from
locally based. Many existing national payment systems, whilst
developing into a full-blown success story. These mixed mar-
often efficient at a domestic level, are fragmented and fail to
ket results have led to persistent resistance to the emergence
satisfy E.U. commission requirements for SEPA. The array of
of genuine BSPs.
institutions through which payments must currently pass has
122 - The
also been a barrier to efficiency in this market. These institu-
In our experience, this resistance is usually driven by a fear
tions also face a trade-off between processing volume and
among potential players that BSP initiatives might actually
Journal of financial transformation
1 This is a revised version of the article published in International Payments, March 2004
bring about more harm than benefit. The reasons and barriers
required to keep up with industry changes, such as CLS, T+1,
for this are multiple. Firstly, many banks still consider their
Basel II, and the E.U. pricing directive, to name but a few.
technology and operations capabilities as valuable sources of
There are many examples of payments businesses where
competitive differentiation. In the battle for market share they
maintenance and support of existing infrastructure exceeds
are reticent to let go of what are believed to be core assets by
80 percent of the available IT budget. Banks are slowly com-
turning to Business Service Providers (BSPs) as an alternative.
ing to digest these cost implications.
These organizations have failed to recognize that in many cases proprietary infrastructure is not core to their business,
Finally, banks fear that the risks to quality and control that
as it does not sufficiently differentiate their offerings. If any-
are involved might outweigh the benefits of outsourcing.
thing, this somewhat antiquated mindset is increasingly
They are afraid to delegate services that could impact cus-
becoming a competitive albatross. In our view, the persistence
tomer satisfaction, that SLAs in themselves will not ensure
of financial institutions to remain infrastructure owners is
quality and that they will lack the weight to influence deci-
neither realistic nor viable, especially for small players. We
sions and manage the new interface as required. They are
continue to advocate the point of view that, regardless of
afraid that the nature of BSP itself will impact quality and
governance structure, the customer franchise business, prod-
customer satisfaction. Lack of exit options from the out-
uct innovation and delivery, and infrastructure provision will
sourcing model once the decision has been made magnifies
need to be unbundled as the payments landscape evolves in
these fears.
the years to come. The lingering resistance to BSPs is based on traditional bankSecondly, banks have not yet overcome their reticence to
ing perspectives on the dynamics of competitive markets.
work collaboratively with competitors when it comes to out-
However, it is the very nature of these dynamics that have
sourcing non-core activities. Many players are concerned
undergone significant change in recent years, particularly in
about the loss of management control and ownership over
the European payments market. This change has been driven
outsourced activities. However, alternatives exist to the sim-
by regulators with an increasingly European and customer
ple customer/Âprovider model and these allow more flexibility
protection outlook, challenging the historical balance
in terms of governance and control. Indeed the governance
between regulation, competition, and cooperation that has
model of a BSP can offer a co-owned or co-sourced service,
developed in each national market. As such, the change in
and as such can allow many providers to operate in an envi-
regulatory stance will force banks to re-think their payments
ronment of cooperation. These governance options provide a
models, and consequently they compete and cooperate. This,
viable alternative to any dominant-player centered models.
combined with emerging standards, new technologies, and increased competition will drive banks to increasingly ques-
Thirdly, despite increasing cost pressures, the traditional
tion the value of keeping payment processing capabilities in-
financial incentive to explore and develop alternative pay-
house. The emergence of infrastructures that undermine the
ment governance structures has simply not been sufficient to
value of traditional correspondent relationships and, for the
induce wholesale banks to begin the transformation process.
smaller providers, the inability to follow their competitors to
More recently, banks have come to recognize and appreciate
low-cost offshore environments, will compound these argu-
that the costs associated with owning all the business ‘bun-
ments. As such, the relationships between traditional com-
dles’ described above are suboptimal for all but the largest
petitors are evolving toward more collaborative footings
providers. Indeed, organizations are struggling to cope with
when appropriate. The old market paradigms are slowly fad-
the overwhelming maintenance and development costs
ing away as value propositions are being redefined around
123
customers, cost pressures are mounting, and issues around quality standards of external partners are being redefined.
ciencies and cost reductions by managing routine business functions — is yet another example. iPSL now handles nearly 70 percent of all checks processed in the United Kingdom and
Change is in the air
is truly transforming business operations, all of which demon-
Against this backdrop, we see the emergence of three pay-
strates that a number of banks can collaborate successfully
ment BSP models in the not too distant horizon. The common
with a large technology partner.
theme across all three of these models is customer-centricity, and if implemented correctly they will have the potential to
Infrastructure-centric model — The emergence of new pay-
liberate a dimension of value beyond cost benefits and equity
ment infrastructure platform players — EBA Step 2/3,
participation. Accelerated top-line revenue growth constitutes
VisaDirect, and domestic players with international ambitions
a third dimension of value, and is a direct outcome of the
— is demonstrating a trend toward making payments both
marked increase in quality that will characterize winning BSP
more regionally oriented and customer focused. Market initia-
models of the future. Higher quality standards will translate
tives seeking to drive common messaging standards — IBAN &
into higher degrees of customer loyalty, and ultimately cus-
BIC, MT103, RosettaNet, and TWIST — are becoming more
tomer-spend, which will be achieved through improved spe-
prevalent, driving automation, and allowing infrastructure pro-
cialization and managerial focus.
viders to get closer to the end customer. Additionally, many incumbents are increasingly enhancing their market proposi-
Bank-centric model (FI insourcing) — Many banks are
tions to offer value-added services, both advisory and process-
acknowledging the benefits of focusing on scale through cen-
ing.
tralization, shared services, and efficiency service offerings. We see this in the current market with both Citibank and ABN
Developing a framework for success
Amro augmenting their respective payments service delivery
Newly established business strategies are shifting banks from
to large corporate customers by creating utilities, access por-
‘infrastructure compliance’ to ‘business enhancement’ and
tals, and support for exchange and marketplaces. Deutsche
driving the realignment of bank operating models to support
Bank is also leading the development of full-service distribu-
this transformation. Higher service levels and cost savings are
tion channels, leveraging their strong retail and wholesale
being achieved through cohesive sourcing strategies which
client bases, as well as their ability to meet the demands for
support a bank’s operating model through the utilization of
unified services across products and markets. These strate-
outsourcing and offshoring. The framework for success thus
gies are allowing banks to solidify their market positions as
lies in developing the case for a BSP solution that demon-
payments systems become increasingly interconnected.
strates substantial long-term benefit while simultaneously being viable in the short- to medium-term. The benefit must
Joint venture model — Market forces are challenging the
outweigh the implementation complexity and any related
traditional payment system continuum. Non-traditional pro-
down side from social and system upheaval.
viders are challenging incumbent systems and payment mar-
124 - The
kets are increasingly responsive to payment provider models
Thus, beyond a compelling financial model, non-captive BSPs
that combine the strengths of world-class banks with global
should additionally be designed from the outset to price ‘at
solutions and technology providers. The partial success of
market’ rather than be a cost-transfer mechanism for incum-
AMS’s PROPONIX offering is a testament to this fact. The suc-
bent institutions, have a governance and consequently pric-
cess of iPSL — the U.K.-based joint venture between Unisys,
ing structure that support open and ubiquitous usage, and
Barclay’s, Lloyds TSB, and HSBC, which seeks to create effi-
from the outset have credibility in terms of capabilities and
Journal of financial transformation
longevity.
Even with these criteria fully satisfied there is a final hurdle. The complexity of providing a payment BSP lies in the complexity of the broader service proposition. Whilst the underlying payment processes are commodity processes, the customer service layer and supporting transaction conditions can create significant complexity. This complexity is, however, at the heart of what customers value and therefore formulates an important dimension on which banks can differentiate themselves from their competitors. BSP providers must be able to either handle these complexities or allow their customer banks to handle them in more efficient ways. The dynamics of the payments market in Europe are evolving at increasingly rapid rates. Customer, competitive, technical, and regulatory pressures are forcing payment providers to re-examine the basis on which they design and ultimately deliver customer propositions. Traditional forms of governance and structure are being replaced with innovative arrangements, often involving either competitors turned collaborators or non-traditional payments players, or even both. Standing still and watching these developments from the sidelines can no longer be a viable strategy for European payments providers. Understanding the nature of the changing game and positioning for success on the basis of a disaggregated value-chain has for many become the only feasible means for long-term survival in the European payments space.
125
Future
Network-based payments and e-settlement
Harry Leinonen Adviser to the Board, Financial Markets, Bank of Finland1
Abstract In order to increase the efficiency of payment systems the opportunities of new technologies need to be employed. Integration of customers ICT systems and those of the banks need to be improved in particular. Internet technology provides new possibilities for direct end-to-end contacts between all participants in the payment processing chain. Making an e-payment could be made as easy as sending an e-mail. This article describes the steps and changes that are, in general, necessary in order to transform national paper-based payment conventions into international network-based solutions. We will need an international account addressing system and an open network between service providers to transport wellstandardized payment messages, including the inter-bank settlement.
1 The views expressed are those of the author and do not necessarily reflect the views of the Bank of Finland.
127
Network-based payments and e-settlement
Decentralized infrastructure
plish final inter-bank settlement in central bank money in real-
The network environment is decentralized and telecommuni-
time for end-to-end payments in straight-through-processing
cation builds bridges between the different independent but
(STP) mode. It should be seen as one essential part of the
interoperable entities in the network. With regards to pay-
whole payment (credit transfer) circle (Figure 3).
ments, these bridges must transfer payment messages as well as inter-bank settlements. This is the main difference with
In Figure 3, (1) the payer receives a bill or other instruction
other messaging systems. In a decentralized network-based
from the beneficiary concerning a payment to be made. (2)
environment the settlement method also needs to be decen-
The payer then sends the instruction to his bank for process-
tralized in order to be efficient. In this environment, the settle-
ing and routing to the beneficiary’s bank. (3) This inter-bank
ment method will need to entail immediate finality between all
leg includes e-settlement, so that the beneficiary’s bank
the different participating service-providing institutions
receives both the payment message and the final settlement.
(mainly banks). The new e-settlement model introduces new
(4) The beneficiary’s bank can then inform the beneficiary as
automated and electronic possibilities for inter-bank settle-
to the incoming/final payment.
ments. E-settlement is a proposed new settlement method for the next generation of payment systems, which can nonetheless be employed now in current payment systems. The inter-bank settlement process itself will have to be a wellintegrated part of the payment process, with end-to-end control from sending to receiving bank. In order to support rapid payment transfers, the settlement method must also support real-time processing. An efficient settlement system also supports continuous reconciliation for immediate error detection. It is also important that most of the security and control features be built into the system, to enable immediate reaction. The e-settlement process is designed to support and accom-
Bank 1 Payment/account server
Payment
Payment
TCP/IP Internetnetwork
Bank 2 Payment/account server
Payment
Payment Bank 3 Payment/account server
Bank n Payment/account server
Figure 1: Direct interbank communication in a network-based infrastructure
128 - The
Journal of financial transformation
Figure 2a: Payments can be made by transferring funds by addressing directly the receiving account in a common account number space.
Figure 2b: The common account number space is divided into sub-spaces belonging to service providers which are connected via interoperable system bridges.
2 See for IBAN-information Thomson Financial Publishing webpage: www.tfp.com/ payment.shtml and the European Committee for Banking Standards (ECBS) webpage: www.ecbs.org
Network-based payments and e-settlement
Current payments clearing conventions, although they now
the address of any account in the system. This could be com-
use electronic automation, have evolved out of paper-based
pared to the address of mailboxes on the Internet or interna-
physical processing and transportation. In order to further
tional GSM telephone numbers for routing SMS messages
increase efficiency and improve services, the payment systems
internationally (Figure 2a). An international account number
need to be reengineered. The main drivers for change will be:
(IBAN) seems to have become the preferred option. If all banks implement IBANs properly, the customer will only need
■ The possibilities of modern network-based real-time pro-
cessing. ■ The need for a clear and universal account number con-
vention. ■ The benefits of a direct decentralized inter-bank settle-
ment facility, e-settlement.
to state the correct IBAN and the payment will be routed to the right account. This will require the systems to have search tables from which the right BIC, bank name, network address, etc., can be found based on the IBAN. These kinds of modules and search tables are emerging2. Without a universal account number-space, efficient cross-border STP cannot be reached. Interoperable bridges are needed to route payments between
The new technology of telecommunications, mainly Internet
the accounts of different service providers (Figure 2b).
technology (TCP/IP) based networks, security facilities like public key infrastructure (PKI) and hardware secure modules (chip cards or secure PC cards), real-time processing, low cost and efficient server hardware, and many others factors will be the basis for this reengineering.
Connecting networks Today everybody is connected via networks. Using Internet and e-mail we can send different kinds of messages, in virtually real-time, to most of the bank employees in the world. However, we lack these possibilities for payment messages. By extending the decentralized network concept of the Internet to payments using secure and dedicated TCP/IP networks like SWIFTnet, we can build a new end-to-end and STP payment process without any intervening clearing and sorting centers. The network sorts the payments by transporting them to the given network address in the same way that an e-mail is routed to the given network server. In the e-mail system banks
Payer’s bank
3
have e-mail/mailbox servers. In a network based payment systems banks will need payment/account servers, otherwise the concept is very similar.
International account number (IBAN)
Including e-settlement
Beneficiary’s bank
Bank-to-bank
4
Customer-to-bank
E-banking
2
Bank-to-customer
Statements receipts
Customer-to-customer
1
Payer
Beneficiary
E-billing
In order to route payment messages efficiently, in networkbased solutions as well as in any traditional system, we need a universal account number convention that clearly indicates
Figure 3: E-settlement is part of the credit transfer circle, which provides efficient electronic communications between participants in a payment.
129
Network-based payments and e-settlement
Inter-bank network
direct confirmation of their payments. A bank that is often
A dedicated inter-bank network is needed to link together all
obliged to inform its customers that payments are queued,
participating banks and the central bank, for the purpose of
that it is waiting for liquidity, will lose customers. In the real-
processing payments. The inter-bank payment network is the
time environment, customers expect direct delivery.
essential part of a network-based payment system. All banks can address each other directly and send payments to each
Still, the e-settlement module could contain basic queuing
other without a centralized processing and routing site. This is
facilities for situations in which the available liquidity is not
the essential new paradigm introduced by Internet communi-
sufficient or customers are willing to accept delays. These
cations (TCP/IP-networks). All participants can operate inde-
would be decentralized queues, designed for different levels
pendently; they only need enough networking capacity to
of complexity. Bilateral netting could be accomplished in the
meet their own needs. System administration is needed only
distributed e-settlement system through bilateral netting
for administration purposes, not, say, for payment processing.
requests to check whether there are transactions also queued
The new SWIFTnet network, introduced by SWIFT, is one that
at the other end. Multilateral netting requires a centralized
can support direct communications between all participants.
netting agent. Different types of netting and advanced liquid-
There are also national dedicated payment networks with the
ity saving features would complicate the system. It is advis-
same capability, such as the inter-bank network Pankkiverkko
able to keep the basic system very simple; possible add-on
in Finland.
services should be provided separately.
Immediate liquidity
High security and availability
For settlement purposes, banks need liquidity. Liquidity is
The system's security features must be carefully designed.
transferred by the central bank to the system (i.e. e-settle-
The settlement balance and all security keys need to be in
ment modules) at the start of the day. It can be increased
tamper-resistant environments and all the encryption algo-
during the day by the central bank via liquidity transfers or
rithms must be highly reliable. There should be no possibility
payments to the banks. At the end of the day, liquidity is trans-
of system intrusion, and any type of hacking should be imme-
ferred back to the central bank. The liquidity in the settlement
diately detectable. The system will be closed, with settlement
modules is thus composed of positive balances of central bank
money circulating among a limited number of trustworthy
money, originally in the traditional form of reserve deposits,
users. The system will include automated reconciliation at
intraday credits, etc., but transferred from the centralized
end-of-day, from time to time during the day, and in connec-
system in the morning to distributed e-settlement modules to
tion with each transaction. In a network-based environment,
be employed during the day in the e-settlement system. In the
all parts, centralized and distributed, must be well secured. A
evening the liquidity will be transferred back to the centralized
digital/electronic version of the four-eyes principle has to be
accounts for overnight bookings. When inter-bank payment
implemented.
systems provide at some point in the future a 24hour/7day service, as has been predicted, the e-settlement modules
Two different controlling programs and two different encrypt-
could start to run continuously by only reporting the balance
ed and tamper-resistant storage devices protect the money
at the turn of accounting day.
balances. The separate controlling functions are monitoring
In a true real-time environment, there is generally little scope
development teams that have all the necessary programming
for the various types of liquidity saving features, based on
code in their possession. In case of intrusion attempts, all
delaying or queuing of payments. Customers are waiting for
critical information will be destroyed at that site and there will
each other so that there are no individual IT developers or
130
Network-based payments and e-settlement
This credit-push/credit-transfer type of payment is the most
graphic technology (e.g., PKI). These stamps are produced and
convenient and efficient in the network real-time world. It has
decoded by e-settlement modules situated close to banks’
fewer processing and transportation legs than electronic
payment systems.
credit/debit card payments, direct debits, or electronic checks. In credit transfers, the payer’s bank identifies its customer,
The e-settlement modules are tamper-resistant devices pro-
checks the payment instruction, and debits the payer’s
vided by central banks to each bank. These are closely inte-
account; the beneficiary’s bank checks the settlement and
grated with banks’ payment systems, for example, directly
credits the beneficiary’s account. In the future real-time world,
integrated with the SWIFTnet access platform (CBT). This
payments will be processed within seconds in the same way as
makes settlement transfers a highly automated part of pay-
e-mail and SMS-messages are now processed.
ment processing. Integrating the new settlement process will be quite straightforward, given that it will be done on the
An e-settlement solution
access platform (e.g., SWIFT CBT) level. In traditional RTGS
The e-settlement solution should be seen as part of the future
systems, banks’ settlement accounts are located in the central-
payment infrastructure that will support an increase in e-com-
ized RTGS system. In the e-settlement system, each bank’s
merce via the Internet, real-time security and money market
settlement account is distributed to the bank’s own processing
deals and transfers, mobile payments (currently GSM-based
site in a central bank controlled e-settlement module. This
but soon UMTS-based), and cross-border volumes.
module should be regarded as a completely automated central bank branch serving one customer with one account. Each
The payment world (for all kind of payments) will be changing,
bank has access to its own account, as before, but is much
as will all other messaging systems, from slower-paced batch
more closely integrated in a more automated and efficient way.
processing to immediate real-time service, integrated directly with user systems in a global network community.
The distribution of central bank money in electronic format to banks’ payment platforms is the essential feature of the
E-settlement provides a solution that can be integrated into
e-settlement approach. The distributing e-settlement modules
current systems, using a part of the existing infrastructure,
need to be highly secure and need to meet at least the same
and hence it facilitates a gradual change from current struc-
security standards as do traditional RTGS systems. The sys-
tures to new e-based ones. The fundamental idea of e-settle-
tem should also be generally open and independent, to sup-
ment is attachment of a digital e-settlement stamp to the
port the various payment networks used by banks.
current payment messages. The e-settlement stamp is added to the payment message and serves to transfer central bank money from payer’s bank to beneficiary’s. Final settlement is part of the payment message, in the form of an electronic central bank draft or central bank e-cash for inter-bank settlement purposes. The electronic stamp will accompany the payment on its route through the inter-bank payment network to the receiving bank. The electronic stamp can be seen as a modern version of a central bank draft. It is the cover in central bank money of the payment(s) it accompanies. The stamp is protected by very strong and modern crypto-
131
Network-based payments and e-settlement
be a security alert to the system administrator(s).
cess itself. Banks need only invest in low cost equipment. The
High availability must also be ensured in the distributed sys-
also transfer payment flows from centralized processing cen-
tem. In a distributed system, a malfunction will generally
ter systems to more efficient decentralized network-based
affect only one participant at a time and only those payments
communications.
very low transaction costs of e-settlement will enable banks to
to and from that particular participant. In order for the participant to reestablish normal operations quickly, there should
The bottlenecks created by centralized resources will disap-
be back-ups and mirrored devices for all critical components.
pear and even the dependence on critical centralized resourc-
Redundant information in the e-stamps gives the possibility to
es will be dramatically reduced. E-settlement could offer a
parse the information of completely destroyed IT-sites. It also
solution for integrating the euro-zone area payment systems,
supports a direct switch of functionalities and services to
and a multi-currency version could serve an even larger area.
working parts of the network. This makes distributed systems
In order to achieve large-scale benefits via the e-settlement
more robust than traditional centralized systems.
model, the number of participating banks and the payment flows must be sufficiently large.
The benefits The main benefit of e-settlement is that it enables redesign of
The e-settlement approach will also reduce settlement risk,
the whole payment system process in an efficient way, using
because all settlements are done in central bank money with
new network possibilities. It thereby creates the next genera-
immediate finality. The reconciliation and control functions in
tion of payment systems infrastructures and makes the settle-
the system will also reduce the possibility for errors or at least
ment process more efficient. Payment systems will change
speed up the process for finding them. In general decentral-
considerably in the near future due to modern technology and
ized systems are more robust than centralized systems.
it would be an advantage to modernize the settlement conventions at the same time. The decentralized network-based model facilitates direct realtime communication. General standards will result in a STP process. The best example is the current e-mail system and its standards. These are applied worldwide and give a really low cost, efficient, and rapid communication system. The e-mail costs are so low that they are considered to be part of the general overhead; nobody bothers to report them separately. The same type of infrastructure with additional security elements and the settlement function could be developed for the payment sector. The cost-advantage of the e-settlement system is in the low processing costs of adding the e-settlement stamp that enables instant final settlement in central bank money. The extra processing cost of adding the e-settlement stamp will be practically zero. It will be an integral part of the payment pro-
132
3 The results of the prototype project are presented at the Bank of Finland’s web site www.bof.fi/sc/e-settlement and in Leinonen, H., V-M. Lumiala, and R. Sarlin, 2002, “Settlement in modern network-based payment infrastructures – description and prototype of the e-settlement model,” the Bank of Finland publication DP 23.
Further information about developments in payment system infrastructures can be found in Leinonen, H., 2000, “Re-engineering Payment Systems for the E-world,” the Bank of Finland publication DP17, and in Payment Systems Worldwide magazine, Spring and Summer 2000 issues.
Future
Banks’ strategies for payment services: Which role for debit cards?
Francesco Saita Associate Professor, Financial Markets and Institutions Department, Università Bocconi
Abstract Debit cards’ role, either used at ATMs or at merchants’ pointof-sale, in payment services provisions by banks has been studied by many researchers before the development of Internet banking. Yet, by analyzing data on recent evolution of debit cards’ diffusion and usage across major countries, debit cards show interesting evolution patterns. Is there still a role for debit cards in banks’ strategies on payment systems? This paper discusses this problem by analyzing why banks might invest in debit cards’ development and examining which factors in practice may make them very cautious in doing so.
133
Banks’ strategies for payment services: Which role for debit cards?
new channel was slower than anticipated and that its short-
store payments cannot be substituted by payments through
term benefits for banks’ P&Ls had often been overstated. This
the Internet, it is possible to have a debate about whether, and
has also led some banks to consider, with much more caution,
up to what extent, debit cards may still represent a significant
investments in new alternative distribution channels, recon-
opportunity for banks in the payment services arena.
sidering the benefits deriving from the traditional branch networks2. This caution, and sometimes the disenchantment, about new channels also has an impact on the strategies in the payment business and on the willingness to pursue and introduce innovations in the field. In this paper, I aim to discuss the role of debit cards to support a bank’s strategy in the payment services business. The reason why this could be interesting is that while they have been a major driver for change and were at the center of the debate about technological innovation before the birth of telephone and Internet channels, their popularity among bank managers has fallen in recent years. While telephone and especially Internet banking were attracting major attention and by far the largest share of banks’ investments in new distribution channels, debit cards, together with ATMs and POS terminals, only remained in the background. At present, then, after realizing that the e-commerce growth was certainly not as profound as many had thought, and that a substantial part of in-
Cards with a cash function Cards with a debit function Cards with a credit function Cards with an e-money function
2001
2002
2001
2002
2001
2002
2001
2002
Belgium Canada France Germany Italy Japan* Netherlands* Sweden* Switzerland U.K. U.S.A.
1361 Nav 711 1480 429 2554 1608 536 1191 2247 2886
1442 nav 741.5 1443 477 2603 1635 643 1232 2400 2996
1217 Nav 652 1252 404 Nav 1315 542 745 920 888
1306 nav 684 1129 459 Nav 1338 549 789 1004 902
296 1510 Nav 381 345 1820 312 419 452 951 4322
294 1653 Nav 391 375 1919 316 472 454 1066 4355
778 2 5 818 1 Nav 1309 63 485 nav nav
800 2 14 819 5 Nav 1078 55 502 nav nav
*Data on cards with a credit function include cards with a delayed debit function. Figure 1: Number of cards per thousand inhabitants Source: Committee on Payment and Settlement Systems (CPSS), Red Book Statistical Update, March 2004.
134 - The
Journal of financial transformation
1 Bansal, P., 2002, “Keep Your Costs Down,” The Banker, March, 125 2 Bansal, P. 2002, “Multi-media Difficulties,” The Banker, March, 124.
Banks’ strategies for payment services: Which role for debit cards?
Retail payment services have always been important for CAGR of ATMs per country, 1998 2000 2002 1998-2002 CPSS average
648
794
927
9.4%
banks, not only because the supply of payment services has been fundamental in their development in the Middle-Ages, but also because payment services currently represent a business area with a substantial impact on banks’ profitability.
U.S.
677
967
1220
15.9%
U.K.
421
563
690
13.1%
Switzerland
642
675
706
2.4%
Sweden
281
295
297
1.4%
Netherlands
418
435
466
2.8%
over time devoted significant effort and investment in order to
Japan
934
922
895
-1.1%
increase revenues and reduce costs deriving from the pay-
Italy
487
549
683
8.8%
ment business. The development of new channels, such as
Germany
556
580
612
2.4%
ATMs, POS terminals, telephone, and Internet banking, has
France
490
580
637
6.8%
represented a clear opportunity to pursue those objectives.
Canada
778
1037
1272
13.1%
Belgium
564
657
683
4.9%
Figure 2: Number of ATMs per million inhabitants Source: Committee on Payment and Settlement Systems, Red Book Statistical Update, March 2004. Note: CPSS average is a weighted average of the countries included in the Committee on Payment and Settlement Systems study, and also comprises Singapore and Hong Kong.
CAGR of POS terminals per country, 1998 2000 2002 1998-2002
According to a 2002 study by Boston Consulting Group, the payment business accounts for up to 35% of revenues and 40% of costs for banks1. As a consequence, many banks have
At the same time, new channels did not always fulfill the perhaps wishful and optimistic expectations of increased profitability that had been formulated ex ante. After the end of the period of over-enthusiasm about Internet banking, many observers agreed that the development of the Internet as a
CAGR 1998 2000 2002 1998-2002
CPSS average
25.4
28.4
25.6
0.2%
12040
18.4%
U.S.
40.6
45.3
36.7
-2.5%
10976
12128
18.5%
U.K.
31.7
34.5
38.3
4.8%
12533
13691
7.0%
Switzerland
11.3
18
21.5
17.4%
10976
11.3%
Sweden
37.6
36.2
36
-1.1%
11439
8.0%
Netherlands
27.1
29.5
29.7
2.3%
10972
6.5%
Japan
3.1
3.2
2.9
-1.7%
9878
14109
23.8%
Italy
8.5
9.1
10.9
6.4%
7194
5584
6.0%
Germany
17.1
20.4
19.7
3.6%
13848
15620
6.7%
France
15.5
17.9
19.8
6.3%
13024
14231
15737
4.8%
Canada
47.5
48.4
45.6
9121
11346
13136
9.5%
Belgium
15.6
19.8
23
CPSS average
6136
10862
U.S.A.
6157
U.K.
10462
Switzerland
7158
9369
Sweden
8406
9822
Netherlands
8533
9774
Italy
6001
Germany
4423
France
12047
Canada Belgium
Figure 3: Number of POS terminals for million inhabitants Source: Committee on Payment and Settlement Systems, Red Book Statistical Update, March 2004. Note: data for Japan are unavailable
-1.0% 10.2%
Figure 4: Number of transactions at ATMs per inhabitant Source: Committee on Payment and Settlement Systems, Red Book Statistical Update, March 2004.
135
Banks’ strategies for payment services: Which role for debit cards?
U.S. are among the countries with more ATMs per inhabitant CAGR 1998 2000 2002 1998-2002
and have a higher usage rate, variations in the number of transactions per capita are much lower than changes in the
CPSS average
18.3
27.2
39.1
20.9%
number of available terminals. The opposite happens instead
U.S.
20.8
33.8
54
26.9%
for POS transactions. As a result, by combining the data, it is
U.K.
30.3
40.7
51.7
14.3%
Switzerland
14.9
23.9
30.8
19.9%
possible to assess that while ATM productivity in terms of
Sweden
18.1
28.9
57.1
33.3%
Netherlands
37.9
50.3
66.2
15.0%
Italy
3
5.5
9.5
33.4%
Germany
8.2
12.7
16.7
19.5%
France
43
54.3
66.9
11.7%
Canada
44.8
63.7
76.4
14.3%
By observing the data in Figure 6 and 7 there are many inter-
Belgium
29.1
39.8
52.1
15.7%
esting elements that emerge. First of all, the two countries
Figure 5: Number of transactions at POS terminals per inhabitant Source: Committee on Payment and Settlement Systems, Red Book Statistical Update, March 2004.
number of transactions per terminal has generally been decreasing (with the remarkable exception of Switzerland), POS terminals have continued to face a significant growth (Figures 6 and 7).
that experienced the highest growth in ATM transactions per inhabitant, Switzerland and Belgium, did not attain this objective through a remarkable increase in the number of available terminals, but by simply increasing their productivity. This suggests that investment on customer education or appropri-
possible to see that the dispersion across countries has even
ate pricing strategies may be more effective than aggressive
increased between 1998 and 2002, especially due to the strik-
strategies concerning the number of available terminals. It is
ingly high growth rates of ATM terminals in the U.S. and
also worth noting that the countries with the most remarkable
Canada in particular, that have been replicated in Europe only
growth in the number of ATMs per million inhabitants (Canada,
by the United Kingdom. In any case, all countries except Japan
U.K., and the U.S.) not only experienced a substantial decrease
show a clear tendency to an increase in the availability of
in ATM productivity, but in two cases also registered a net
terminals; overall, the number of ATMs in the countries cov-
decrease in the number of transactions per inhabitant.
ered by the Committee on Payment and Settlement Systems report in 2002 was 40% higher than four years before.
Data on POS productivity are different, since all countries
The picture concerning the availability of POS terminals also
actions per inhabitant that productivity increased every-
experienced such a significant growth in the number of transshows clear differences between countries, even if in this
where. There is also a positive correlation between the
case differences seem to be smaller and many European
growth in the terminals network and the growth in the num-
countries appear to be in leading positions. Among those,
ber of transactions per inhabitant (as opposed to the nega-
particularly remarkable is Italy’s compounded annual growth
tive correlation that existed for ATMs). Yet, expanding the
rate of 23.8%, with POS terminals more than doubling
POS network is clearly not the only way to increase debit
between 1998 and 2002.
cards transactions: Italy and Sweden experienced the highest
While the data in Figures 1, 2, and 3 seem to provide a clear
the number of terminals was only around 30 percent. At the
growth in transaction per inhabitant, but Sweden’s growth in
136 - The
overall picture of the countries where cards have a major role,
same time, Italy itself now has a terminal base per inhabitant
it is interesting to look carefully at the data concerning card
higher than all other European countries except France, and
usage (Figures 4 and 5). For ATMs, while Canada, U.K., and the
yet the productivity of its terminals is lower than 25% of the
Journal of financial transformation
Banks’ strategies for payment services: Which role for debit cards?
Transactions per terminal Var% Terminals Trans CAGR actions/ inhabitant 1998- 1998- CAGR 1998 2000 2002 2002 2002 1998-2002
Transactions per terminal Var% Terminals Trans CAGR actions/ inhabitant 1998- 1998- CAGR 1998 2000 2002 2002 2002 1998-2002
Belgium
27,660
30,137
33,675
21.7%
4.9%
10.2%
Belgium
3,190
3,508
3,966
24,3%
9,5%
15,7%
Canada
61,054
46,673
35,849
-41.3%
13.1%
-1.0%
Canada
3,440
4,476
4,855
41,1%
4,8%
14,3%
France
31,633
30,862
31,083
-1.7%
6.8%
6.3%
France
3,569
3,921
4,283
20,0%
6,7%
11,7%
Germany
30,755
35,172
32,190
4.7%
2.4%
3.6%
Germany
1,854
1,765
2,991
61,3%
6,0%
19,5%
Italy
17,454
16,576
15,959
-8.6%
8.8%
6.4%
Italy
500
557
673
34,7%
23,8%
33,4%
Japan
3,319
3,471
3,240
-2.4%
-1.1%
-1.7%
Japan
nav
nav
nav
nav
Netherlands
64,833
67,816
63,734
-1.7%
2.8%
2.3%
Netherlands
4,442
5,146
6,034
Sweden
133,808 122,712
121,212
-9.4%
1.4%
-1.1%
Sweden
2,153
2,942
Switzerland
17,601
26,667
30,453
73.0%
2.4%
17.4%
Switzerland
2,082
2,551
U.K.
75,297
61,279
55,507
-26.3%
13.1%
4.8%
U.K.
2,896
U.S.A.
59,970
46,846
30,082
-49.8%
15.9%
-2.5%
U.S.A.
35,768
27,616
-29.5%
9.4%
0.2%
CPSS average 39,198
Figure 6: Productivity of ATM terminals Source: Elaboration on data from CPSS, Red Book Statistical Update, March 2004.
nav
nav
35,8%
6,5%
15,0%
4,992
131,8%
8,0%
33,3%
2,806
34,8%
11,3%
19,9%
3,247
3,776
30,4%
7,0%
14,3%
3,378
3,079
4,453
31,8%
18,5%
26,9%
CPSS average 2,982
2,504
3,248
8,9%
18,4%
20,9%
Figure 7: Productivity of POS terminals Source: Elaboration on data from CPSS, Red Book Statistical Update, March 2004.
Of course, when discussing this topic we must consider that
alternative competing ATM networks that may require the
the diffusion and usage of cards is markedly different across
customer to hold multiple debit cards, or the development of
countries, and so is obviously the potential for improvement.
a unique national ATM network), and partly to the character-
For this reason, we will start by analyzing data concerning the
istics of retail distribution network of goods and non-financial
penetration of cards, the availability of terminals, and cards’
services (since the level of concentration, the importance of
usage across major countries. We will then discuss the poten-
big distribution networks, and the average size of retailers’
tial benefits banks may expect from new efforts in cards’ dif-
facilities may impact the speed of adoption of new means of
fusion and examine the obstacles that might prevent banks
payment). The overall recent picture of cards’ diffusion is
from investing significantly in the debit card business.
reported in Figure 1, where it is possible to note that while there is a huge difference, as far as credit cards are con-
Debit cards diffusion and usage across countries: Key data and trends
cerned, between Japan and Anglo-Saxon countries on one
The diffusion of cards across different countries is, as one may
ation is somewhat more mixed as far as debit cards are con-
expect, significantly different, due to a number of reasons that
cerned, where countries such as Netherlands, Belgium and
are linked partly to structural characteristics of the financial
Germany have a clear leading position.
side and remaining European countries on the other, the situ-
system (i.e. different households’ attitude towards financing purchases with debt that may help explain differences in
As for cards’ diffusion, even the number of terminals available
credit cards diffusion and usage across countries), partly to
for card usage is markedly different across markets. Figure 2
the history of the banking system (i.e. the development of
reports the number of ATMs per million inhabitants, and it is
137
Banks’ strategies for payment services: Which role for debit cards?
may complete autonomously through an ATM instead of
sion and usage, a realistic analysis must also carefully con-
entering the branch could further reduce the workload for
sider the obstacles and the doubts that may prevent banks
branches. Debit card and ATM supporters argue that while a
from doing that, or, at least, from investing at the rate tech-
significant cost saving may already have been attained by
nology suppliers and innovation supporters would like. The
most banks, a further development of self-service terminals
reasons behind banks’ caution are various, sometimes
even inside the branch may be fundamental to keep the over-
extremely well grounded and sometimes purely psychologi-
all costs of a bank’s distribution network sufficiently low and
cal. However, they must clearly be considered if one wants to
to let the branch be restructured so as to maximize its selling
realistically evaluate the potential for further debit card
potential. On the other hand, however, multifunction ATMs
expansion. For clarity, we can group the causes for resistance
could be used to deliver other non-banking services (i.e. pur-
into two different areas, the uncertainty about the attainable
chasing tickets for entertainment or sport events) that may
return on investments on debit cards in particular and on the
produce extra revenues for the bank and help amortizing the
payment business in general, and the (at least perceived)
costs of its ATM distribution network. In some cases, develop-
unsatisfying performance of recent investments on new tech-
ing these services may also provide customers an incentive to
nological channels, such as Internet and PC banking.
get more familiar with ATMs. In Italy, for example, banks started offering customers the opportunity to conveniently
First of all, it is very difficult to assess the return on invest-
recharge the prepaid cards for mobile phones through ATMs.
ment from alternative channels precisely. While for instance
Considering the extremely high diffusion of cellular phones in
the extra revenues generated by increased debit card usage
Italy, this has been a significant source for new income and
at POS terminals may be easy to assess, the potential cost
the opportunity to make debit cards a normal instrument for
savings derived from a reduction in the amount of cash being
a large share of customers that had until then proved to be
handled by the bank are clearly much more complex to quan-
resistant to all the efforts devoted to increase card usage.
tify. Consequently, while it may be easy to describe in theory that this may be an advantage a bank should consider when
Developments in e-commerce could also help to increase the
deciding whether and how to promote higher card usage
usage of debit card, if they can also become a safe means of
rates, the uncertainty about the savings that could actually
payment for purchases through the net. This could become
be attained is a clear obstacle. A bank would like to be able to
possible if and when low-cost card readers become available,
fully evaluate the overall economic consequences of either an
so that the buyers can use PIN-based payments confirmation,
aggressive pricing strategy or a marketing campaign among
rather than having to send their card details through the net.
customers before undertaking them, and the difficulties in
Of course similar card readers could be designed for both
understanding the impact on internal costs is from this view-
credit and debit cards, which makes determination of their
point a major problem. This problem is common to other
impact on debit card usage less certain. This is not, however,
channels as well, a typical example being the Internet where
something that may attract banks’ interest in the very short
even a clear assessment of incremental revenues may be dif-
term, since the development of cheap readers is a technologi-
ficult, since it is not always clear whether a product pur-
cal precondition for any effort in this direction.
chased on-line is a real new purchase or simply a substitute
What may prevent banks from investing in debit cards?
more traditional channels.
While there are many good reasons why we should expect
The problem of evaluating real attainable cost savings is
banks to spend time and effort in increasing debit card diffu-
equally important when critically assessing the benefits of
3 Fees can also be less evident in cases when the customer is charged by the bank also on the basis of the number of transactions on his bank account, or the number of transactions over a certain threshold. In this case a customer that uses his debit card five or six times at POS terminals instead of withdrawing cash once and
paying the merchants in cash is clearly generating a higher amount of fees for the bank, even if he probably does not perceive the debit card as costly. 4 Atkins, W., 2004, “Is Debit the New Cash?” The Banker, June, 110-111
for a purchase that the customer would have made through
138
Banks’ strategies for payment services: Which role for debit cards?
is not explained by a lower number of debit cards either (from
Why should banks invest to develop debit cards further?
Figure 1 it is possible to see that for the number of debit cards
The heterogeneous picture emerging suggests that in some
per thousand inhabitants it is close to countries such as
countries there are still large margins for improvement in
productivity of the second worst country reported; the result
Sweden and France with a completely different number of
terms of debit cards’ diffusion and usage. While the potential
transactions per terminal). While differences in terminal pro-
for debit cards’ further development is testified by still high
ductivity may depend also on the structure of retailers’ distri-
growth rates in the number of transactions and is obviously
bution network across different countries, it is quite clear
stressed by new technology suppliers or by networks such as
that efforts devoted to increase card usage rather than sim-
MasterCard or VISA, it is important to discuss which benefits
ply increase the number of cards in circulation or the number
banks may actually expect from further investments on debit
of available terminals may be important. Of course, providing
cards.
incentives to customers to increase card usage by either adopting aggressive pricing strategies or developing cam-
As far as potential benefits are concerned, we could consider
paigns to persuade them to gradually change their behavior
the benefits deriving from greater usage at the point-of-sales
requires a significant commitment. I will now briefly discuss
terminals. In this case, benefits for the bank may derive from
why we should expect banks to seriously consider investing in
the fees that can be charged to customers, if a fee is charged,
this direction and which, at the same time, are the reasons
and merchants. In any case, direct charges on debit cards
why they may still be very cautious and reluctant to do that.
usage would clearly be counterproductive on the number of transactions3 when revenues from merchants depend on the market share the bank has among merchants only (so that, for instance, a bank with a large retail customer base but few merchants has from this viewpoint little incentive to invest to convince its customers to increase card usage). A greater use of debit cards at POS terminals may also imply a reduction in the costs a bank faces for handling cash, and moving it between branches and automated tellers in order to make it available to its customers. Some analysts suggest that if merchants’ fees were based on the real costs the banks bears, the threshold where debit cards would become a cheaper payment type for merchants than cash would drop to a level as low as U.S.$124. While there may be uncertainty about the exact cost estimates that may help identify the threshold precisely, pricing is undoubtedly a critical issue for providing banks, merchants, and customers the right incentives to substitute cash with more efficient means of payment. Apart from debit card usage at POS terminal, banks may be willing to develop usage at automated tellers also, for different reasons. Increasing the number of cash withdrawals, request for information, and bill payments that customers
5 Fernandez Caro, S., 2002, “More Effort Goes into Branches,” The Banker, August
139
Banks’ strategies for payment services: Which role for debit cards?
adequately exploiting this unutilized potential first, rather
important tool for a bank’s strategy in the retail banking busi-
than increasing the capacity of other alternative channels
ness. Yet, banks’ investments in the field may be limited by the
such as, for instance, ATMs. This is reinforced by the fact that
difficulties in clearly understanding channel profitability and
PC and Internet banking may, firstly, provide even greater cost
the actual cost savings that may derive from increased debit
savings for low value-added transactions, such as a balance enquiry or simple bank transfers6, and secondly, offer to the
card usage. If profitability is not completely clear and at the same time debit cards are no longer perceived as a source of
bank cheaper and richer information on the customer’s trans-
differentiation, since innovation is mainly driven by technolo-
actions and interests, which could lead to better selling oppor-
gy producers and international card networks that obviously
tunities. Suggesting new products through the website and
try to spread innovation as fast as possible across banks, a
providing fast and adequate information on their characteris-
certain caution is understandable. Going forward, banks
tics is possible through Internet and PC banking, but it is basi-
should not only limit their investments to hardware and tech-
cally impossible through ATMs. The payoff banks may expect
nological support, they should also try and influence custom-
from trying to exploit the potential of their previous invest-
ers’ behavior through pricing and education.
ments on Internet technology, even if of course the need for upgrades is even an issue here, is therefore substantially different from the one that may be derived by older technological channels. The second implication of the poor result of Internet investments has probably been to increase top managers’ risk aversion when evaluating investments in distribution technology in general, especially considering the difficulties faced in properly measuring, even ex post, returns on investment. It may be hard for the head of a retail banking division to persuade the chief executive officer that despite recent unfortunate results, investing in new channels may be a winning choice for the bank. It is more likely instead to assume that a bank may react to unsatisfying Internet investments returns by reducing investments in the retail banking division in favor of others, and concentrate retail banking investments on restructuring traditional bank branches that have been proven to be far more important to customers than many Internet supporters would have predicted a few years ago.
Conclusions Debit cards that had been for a long time considered to be a key innovation in banks’ distribution networks — prior to the advent of telephone, Internet, and PC banking channels — continue to experience significant growth, especially as far as usage at POS terminals is concerned, and still represent an
140 - The
Journal of financial transformation
6 Bielski, L., 2003, “Hard to Get the Online Habit,” ABA Banking Journal, February, 79-86
Banks’ strategies for payment services: Which role for debit cards?
greater debit card usage through ATMs. In this case, one
Doubts concerning the return on investment from developing
major issue is represented by the fact that while activity-
ATMs may not be limited to only basic financial services, they
based costing techniques may produce an estimate of the
could also be extended to functions that more sophisticated
cost of all types of transactions for a bank depending on the
ATMs may perform and the delivery of non-financial services.
channel the customer uses, understanding clearly which
In fact, extended functionality does not imply extra revenues
parts of those costs can be actually reduced or eliminated by
but clearly extra fixed costs due to the upgrade of existing
migrating transactions towards ATMs or other channels is not
terminals, which may be non-trivial especially for larger ATM
simple. Major savings could derive in theory from reduction
networks. In practice, it is realistic to assume that functional-
in staff costs, especially for cashiers. Yet, while it may be
ity extensions may occur only in those phases that a bank is
simple to use self-service channels to reduce the number of
forced to invest in substituting older-generation ATMs due to
cashiers in each branch to one only, dedicated to complex
the clear obsolescence of the network.
transactions or to resistant customers, pursuing further cost savings by eliminating the cashier completely is not equally
The need to undertake maintenance and upgrading of existing
easy. The risk is, therefore, that cost savings may remain
network is just an example of the kinds of investment that a
largely theoretical, due to the clear discontinuity of the cost
bank may face simply in order to maintain the same service
function for each branch. While the possibility that some
quality as its competitors, and that may certainly influence
banks may still have an incentive to promote ATM usage cannot obviously be ruled out5, potential advantages may be
their willingness to invest further in more ambitious card
higher especially for those banks that had not restructured
relevant and important technological change, such as the
usage development programs. From this viewpoint even a
their branches yet, and maintain branches with a higher than
migration of debit cards from magnetic stripe to chip card
average size.
technology, while opening up new opportunities for safer and greater card usage, will necessitate significant investments in
Customer education efforts devoted to increasing larger ATM
upgrading merchants’ terminals. Again, immediate costs may
use may be particularly difficult to justify, especially consid-
be clearer and more evident to both banks and merchants
ering that similar efforts devoted to increase cards’ usage at
than future returns, or the possibility of becoming late movers
point-of-sales or increase Internet banking may be more pro-
in such a field.
ductive. While a greater development and usage of POS terminals might reduce the need for cash for small transactions
A second source of resistance to investment on debit card
and therefore make cash withdrawals more infrequent, which
may be associated with recent experiences of investments in
might increase the attraction of investments aimed at
Internet banking. In many countries there had been a clear
increasing POS usage, other channels, such as telephone or
overstatement of the growth potential of Internet and PC
PC/Internet banking, allow customers to get information
banking channels. Many banks had, therefore, invested signifi-
about their current account as well as undertake other pay-
cant amounts of money (perhaps partially for a sincere trust
ment transactions, such as bank transfer orders. When com-
on Internet potentiality, and partially simply for an uncon-
pared to an ATM, these channels also offer the bank a better
scious ‘me-too’ strategy) that did not generate adequate
opportunity to sell other products when the customer con-
returns. This fact has had at least two consequences that may
tacts the bank for a payment transaction. As a result, it is
be relevant for our analysis.
likely to think that customer education efforts might be devoted to supporting the development and usage of tele-
The first consequence is that if there is overcapacity with PC
phone or PC/Internet banking rather than the usual ATM.
and Internet banking, it looks natural to concentrate efforts on
141
Future
Delivering migrant workers’ remittances
Roger Ballard Director, Center for Applied South Asian Studies, University of Manchester
Abstract As globalization has led to ever higher levels of labor mobility,
find that they are up against very effective competition: the
so the volume of funds remitted to their families by workers
Informal Value Transfer Systems which many migrant groups
employed in countries far distant from their homes has
have themselves developed to meet this challenge. In a post
increased by leaps and bounds. The total volume of such
9/11 context, there has been much fevered discussion about
transfers currently amounts to over U.S.$100 billion per
the threat which such ‘underground banking’ networks alleg-
annum, the greater part of which flows from economically
edly present to the stability of the global order. In an effort to
advanced regions in the West and North to developing coun-
replace ill-informed speculation with empirically grounded
tries in the East and South. Delivering those funds swiftly,
detail, this paper demonstrates the extent to which such
reliably, and cheaply to relatively remote destinations opens
informal networks have developed in such a way that they
up new opportunities for the financial services industry, but
now provide a highly efficient response to the logistical chal-
also represents a major logistical challenge.
lenge of delivering migrant remittances. In doing so it also sets the scene for a more informed debate about how IVTS
Although many banks and money transfer agencies have
networks might best be regulated.
recently begun to consider how they can make the most of these opportunities, those that seek to enter this field soon
143
Delivering migrant workers’ remittances
population of migrant workers’ consequent need to find a
more importantly, questions began to be asked about the
safe, speedy, and reliable means of transferring their savings
extent to which these IVTS networks were being used to pro-
back to their kinsfolk, especially when, as was very often the case, they lived in relatively remote rural areas. Hence just as
vide a cover for laundering the profits of drug smuggling. And although initial reports5 were rather skeptical about the
the migrants organized their own means of entry into destina-
extent to which IVTS was actually being used as a vehicle for
tion labor markets, which in many cases involve risking their
criminal malfeasance, rather than a convenient means of
lives, they also devised their own means of sending savings
facilitating wholly legitimate remittance transfer, the whole
back home. Whilst the physical transfer of currency notes was
debate was transformed in the aftermath of the events of 9/11.
the most obvious way of doing so, long-distance migrants, and
Even though the greater part of the 9/11 operation was
especially those whose residence rights were uncertain, were
financed by wire transfers into personal accounts opened at
unable to make home visits with any frequency. But as they
the Suntrust Bank in Venice, Florida6, a mass of luridly-written
soon discovered, the formally constituted financial sector was
articles soon appeared in the press suggesting that Hawala/
ill-equipped to meet their needs. Over and above delivery
IVTS networks were in fact a wholly unregulated form of
problems, the commissions charged by banks, let alone by
‘underground banking’ which provided drugs smugglers and
specialist money-transmission agencies, such as Western
terrorists a vehicle for financing their evil deeds. Hence,
Union, consumed an alarmingly significant portion of their
despite the absence of any substantial body of evidence sup-
hard-earned savings. Hence wherever groups of labor-
porting such accusations, governments throughout the world
migrants established ethnic colonies of any size they invari-
have found themselves under ever increasing pressure, for
ably went on to develop their own self-constructed strategies
the most part articulated by the U.S. Treasury, to introduce
of informal value transmission.
regulatory measures which could in effect put all such informal operations out of business.
As the years have passed, most of these informal alternatives
144
have become steadily more sophisticated, aided by advance-
However, as everyone within the financial services industry is
ments in communication technology, particularly in terms of
now very well aware, draconian initiatives whose central
the development of fax and then the Internet, which have
objectives were to counter both money laundering and terror-
made it possible to organize larger-scale value transfers with
ist financing were by no means limited to the informal sector.
speed, accuracy, and with high levels of reliability. Hence
Suddenly virtually every agency engaged in any kind of finan-
whilst these informal transfer systems went largely unnoticed
cial transactions on behalf of personal customers found them-
during their initial years of development, with the turn of the
selves required to introduce elaborate Know Your Customer
century they suddenly became the focus of much higher lev-
procedures, and also under a duty, backed up by severe crimi-
els of institutional attention. There were two reasons for this
nal sanctions, to submit ‘suspicious transaction reports’, espe-
sea change. In the first place once millions and then tens of
cially in the case of cash transactions in excess of U.S.$10,000.
millions of dollars began to be transmitted through these
Whether this new regulatory environment is having any effect
Informal Value Transfer Systems (IVTS), they began to be suf-
on the problems it is designed to confront remains to be seen.
ficiently large as to make blips on institutional radar screens.
What is undoubtedly the case, however, is that as a conse-
Hence a number of major banks, led by those based in the U.S.
quence of these initiatives it is not just IVTS networks which
with an outreach into Latin American markets, began to
have found themselves under pressure. Financial institutions
explore whether it made commercial sense to make a more
of all kinds now find themselves faced with the necessarily
active response to the potentially-profitable business of facili-
expensive task of gathering and recording a great deal more
tating remittance transfers. Secondly, and ultimately even
personal data about their clients and their transactions than
1 Ratha, D. “Workers’ Remittances: An Important and Stable Source of External Development Finance”. Global Development Finance Report. World Bank, Washington D.C. (2003) 2 House of Commons International Development Committee. “Migration and Development: How to make migration work for poverty reduction” HC79-1. The Stationary Office, London (2004)
3 An overview of the growth and qualitative characteristics of Britain’s South Asian population can be found in Ballard, R (Ed). “Desh Pardesh: The South Asian Presence in Britain.” C. Hurst and Co., London (1994). 4 Piore, M. J. “Birds of Passage: Migrant Labour and Industrial Societies.” Cambridge University Press, Cambridge (1979).
Delivering migrant workers’ remittances
The growth of migration and remittances in the contemporary world In the context of an ever more globalized labor market, many millions of people are now living and working in locations far
holders — stretching from migrants themselves to their families and countries of origin to the International Development Com munity — have an interest in ensuring that migrant remittances are as swift, reliable, and above all cheap as possible.2 It fol-
distant from their countries of origin, and to which they regu-
lows, therefore, that the recent rapid growth in South-North
larly remit a substantial proportion of their earnings. The
labor migration, and the concomitant growth of remittance
financial flows generated by migrant remittances in this sense
transfers in the reverse directions, presents both a major new
are by now very substantial. Although accurate figures are
challenge and an equally substantial set of novel commercial
extremely hard to come by, it is not unreasonable to suggest
opportunities to the financial services industry. This paper
that the current flow of transnational value-transfers cur-
explores the ways in which those opportunities have so far
rently generated by migrant workers amounts to at least
been met, identifies the kinds of institutional initiatives which
U.S.$100 billion per annum, and perhaps a good deal more.
have emerged to meet them, and on that basis seeks to look forward to the likely course of future developments.
Nevertheless migrant workers are far from being a homogeneous socio-economic category. Although the professionally
Although migrants have always played a major part in provid-
qualified employees of multinational companies now tend to
ing the labor power needed to fuel urban and industrial
display high levels of spatial mobility during the course of
growth, and despite the fact that they continue to be a vital
their careers, such high-fliers only make up a very small pro-
source of menial workers in most post-industrial societies, and
portion of the global stock of migrant workers. Instead the
most especially those with plummeting birth-rates, their
vast majority work at far lower levels in the occupational hier-
arrival has rarely been planned or centrally coordinated.
archy. But although their wages are only a small fraction of
Moreover all efforts to achieve such coordination invariably
those enjoyed by the high fliers, most display a much higher
fail, for one very obvious reason. Migrant workers are not only
propensity to save, so much so that they frequently send a
invariably self-selecting, they also behave as self-motivated
substantial proportion of their relatively meager incomes
entrepreneurs whose principal concern is to advance the interests of themselves and their kinsfolk.3 Migrants have
back to their kinsfolk in Asia, Africa, and Latin America. The
more than a few hundred dollars a month, mere penny pack-
always acted thus. No less than the millions of European migrants who crossed the Atlantic during the 19th and the 20th century, their ever more numerous successors from
ets in banking terms. However it is the volume of such trans-
Asia, North Africa, and Latin America seeking their fortunes in
actions which gives them their current significance, no less
the affluent cities of other continents are pursuing their own
locally than globally. In many countries of the developing
agendas on their own terms. In doing so they not only aimed
size of the resultant transactions, examined one by one, is very small. The vast majority of migrants rarely remit much
world migrant remittances now form the largest single source
to make the most of the income-earning opportunities avail-
of foreign exchange, comfortably exceeding the sum of all forms of development assistance and FDI.1
able at their destination, but were equally committed to increasing the wealth, and hence the status, of their families and their communities back home.4
Not only has the scale of these remittance flows grown very rapidly in recent years, but so, too, has the logistical challenge
Yet although a reverse flow of remittances follows the arrival
of delivering them, not least because so many of their recipi-
of migrant remittances almost as surely as night follows day,
ents live in remote rural areas, far beyond the reach of the
until very recently the formally constituted financial sector
formal banking system. However a growing number of stake-
made no ordered efforts to respond to this rapidly expanding
5 Passas, N., “Informal Value Transfer Systems and Criminal Organizations; a study into so-called underground banking networks.” http://www.minjust.nl:8080/b_ organ/wodc/publications/ivts.pdf (1999). 6 “The 9/11 Commission Report.” Government Printing Office, Washington D.C.
145
Delivering migrant workers’ remittances
U.S.$ 1.8 billion) per year back to Pakistan.7 Apart from a brief
ticipants’ records by accountants and regulators, around
blip when Customs and Excise arrested a series of major
which contemporary Western-style banking systems are rou-
Hawaladars and charged them with money-laundering, the
tinely constructed. Indeed the only external authority to
greater part of these funds were transferred back to Pakistan
whom Hawaladars (the vast majority of whom were Muslims)
through the informally grounded Hawala system, often
were responsible was Allah himself.
through deals facilitated in Dubai. However Pakistanis are not alone in using this system. To my knowledge similar Dubai-
But, although most Hawaladars took, and still take, the treat of
centered transfer networks have also been established by
Divine sanctions seriously, none are so idealistic as to rely on
immigrants from Bangladesh, Pakistan, Afghanistan, Iran, and
that as the principal source of system security. Instead they
Somalia. Such networks also facilitate transfers from immi-
relied on relationships of absolute trust. Not only did the sys-
grants who have established themselves across the length
tem rely on customers having absolute confidence in their
and breadth of the E.U.
Hawaladars’ reliability, the Hawaladars themselves needed to have a similarly comprehensive level of trust in each other’s
These networks did not spring into existence out of the blue.
honesty. This was necessary since due to the distances cov-
Whilst the growth of migrant remittances may have dramati-
ered there were imbalances between the participants for long
cally boosted the scale and scope of contemporary Hawala
periods at a time.
networks, the roots of these operations can be traced back to the ancient system of commercial banking, which grew up to
Because of the need for absolute trust, it is hardly surprising
serve the needs of long-distance traders in the Indian Ocean
to find that each network of corresponding Hawaladars,
region. In their original format Hawala networks emerged to
together with their customer-base, was normally language,
oil the wheels of commercial activity, meeting the needs of
religion, and hence community-specific. From this perspective
long-distance traders who wanted to transport the bullion
Hawala networks are best understood as an Islamic version of
they needed to finance their activities. A deposit made with a
a much wider range of pre-modern forms of long-distance
Hawaladar in one market place (the Gujarati port of Surat, for
value transmission and settlement, to which parallel Hindu
example) could be picked up as a credit from that Hawaladar’s
and Chinese forms of IVTS also emerged during the mediaeval
corresponding partner in Basra, Mogadishu, or Malacca.
period, and likewise remain in operation to this day. Moreover it is quite clear that whilst each such system has its own spe-
As a result of completing such transactions on their custom-
cific set of cultural and religious roots, all are constructed
ers’ behalf, corresponding Hawaladars were constantly shift-
around translocal networks of absolute interpersonal trust.
ing debits and credits between one another. Just as in any
These not only provide a secure channel of communication
other banking system, Hawaladars were prepared to take
between spatially separated financial operators, but are also
these positions because those within any given network
the foundations of a web of binding reciprocities which pro-
trusted their partners to honor each other’s instructions, and
vides, in each case, a bedrock of system security.
above all to make up any net deficiencies as and when a settlement was eventually implemented. Although system secu-
146
Whilst religious and sectarian loyalties provided the basic
rity was as crucial a feature of these operations as it is in any
foundations of such networks, these were, and still are, rou-
other banking system, its foundations were grounded in the
tinely reinforced by ties of kinship and marriage between the
cultural and religious context of the times. Medieval
participants, so generating an even tighter web of mutual
Hawaladars largely eschewed the elaborate clerical and
reciprocities. Hence it is through the operation of such infor-
bureaucratic procedures, such as regular inspection of par-
mally constituted networks, rather than through the imple-
7 The empirical data around which this section of the paper is constructed derives from more than thirty years’ experience of ethnographic research on the growth of the South Asian presence in Britain and its transnational connections, which has in recent years provided a platform for a much more detailed examination of the activities of Hawaladars operating in the U.K. and Dubai, as well as in India and
Pakistan. Between 2000 and 2002 I received a grant from the E.S.R.C. to conduct research on Kinship, Entrepreneurship, and the Transnational Circulation of Assets. A wealth of more detailed material on these issues can be found at http:// www.art.man.ac.uk/CASAS/pages/papers.htm
Delivering migrant workers’ remittances
they had hitherto collected, and even then can often find
ist cash delivery operations based in the formal sector, of
themselves in an uneasy position for fear that they might find
which Western Union is currently the most salient example,
themselves charged with colluding in criminal activity.
regularly charge a commission of around 15% for their services, and even then their delivery-coverage is largely restrict-
A new opportunity for the financial services industry?
ed to major urban centers, especially in the developing world.
Yet despite the costly downside of these new, albeit still much
on their hard-earned wages appears to be quite outrageous,
From the perspective of ill-paid migrant workers a cut of 15%
disputed, regulatory regimes, these post-9/11 developments
especially since the margin charged by operators in the infor-
may, at least on the face of it, have opened up a major new
mal sector is dramatically smaller, often 2% or less. Why, then,
opportunity for the financial services industry. A stable, and
does Western Union find it necessary to charge so much? And
indeed a steadily expanding flow of value transfer to the tune
how, by contrast, do operators in the informal sector manage
of U.S.$100 billion per annum is certainly not an opportunity
to offer a service which is just as reliable and with a much
to be sniffed at, most especially when regulatory authorities
greater spatial reach whilst charging so much less?
around the globe are actively encouraging the formal banking sector to become much more heavily involved in providing
The answer to the first question is obvious enough. Serving
services in this sphere, thereby providing migrant workers
the financial needs of millions of migrant remitters drawn
with an alternative to what the burgeoning AML/CFT industry
from developing countries is an accountant’s nightmare. Given
regards as deeply suspect IVTS and Hawala networks.
that the packet-size of each transaction is small and the delivery address is so often as remote as it is obscure, not only
Yet despite the obvious opportunities that these develop-
must elaborate and expensive arrangements to consolidate
ments have thrown up, those players who have sought to
and then to deconsolidate these small packages at each end
enter this sector of the global foreign exchange market
of the transmission chain be developed, but a process of cur-
appear to be finding the going extremely tough. Devising a
rency exchange, on a wholesale basis of course, must also be
commercially viable response to the specific character of cus-
implemented somewhere along the way. Because of the large
tomer demands in this sphere is proving to be far from easy,
number of the administrative staff required to implement
not least because their competitors in the ‘informal’ sector
these processes when conventional bureaucratic procedures
still appear to occupy a position of clear competitive advan-
are employed, costs are necessarily high. No matter how hard
tage, despite the high levels of official harassment to which
they try to cut down on administrative costs, Western Union
they have recently found themselves subjected.
and its many emerging rivals within the formal sector see little
So just what are the financial and logistical requirements of a
their present levels. No wonder their competitors in the infor-
successful remittance transfer operation? Overall volume is
mal sector are currently having a field day. How, though, are
not a problem. What does cause severe problems is the logis-
they managing to pull it off?
prospect of reducing their commission charges by much from
tics of such operations. Given that individual transfers are mostly in penny packets, most transfers require funds to be
The roots of Hawala
converted into relatively obscure currencies, most senders
In the U.K. the Pakistani community, which is currently around
prefer to make their deposits in cash, and recipients frequently
three quarters of a million strong, is the most active sources
live in rural areas, and hence well beyond the banking frontier.
of migrant remittances, and its members are collectively responsible for sending anywhere between £500 million
In view of these challenges it is hardly surprising that special-
(approximately U.S.$ 900 million) and £1 billion (approximately
147
Delivering migrant workers’ remittances
of current exchange rates to prospective senders, collection of
hours, with no significant organizational fuss and 100% reli-
data from senders about how much is to be delivered where
ability. From the customer’s perspective sending money back
and to whom, and the transmission of delivery data to the
to Mirpur could hardly be more straightforward. All they have
remote destination.
to do is approach their local Pakistani travel agent (the vast
The second is value transmission and settlement, which
or as the agent of a larger Hawala operator based in a city
means that they need to organize availability provision of
which supports a large local Pakistani population), enquire
majority of whom offer such services, either in their own right,
credit at the point of delivery against accumulated debits at
about and agree upon a rate of exchange, prove their identity
the point of deposit, so enabling the organization of subse-
(a requirement of the recently introduced anti-money launder-
quent processes of settlement.
ing regulations), make the agreed payment, and provide the recipient’s name and address in Mirpur. Within 48 hours the
The third element is collection and delivery of currency notes,
sum agreed upon will be ready for collection in Mirpur.
which involves collection and consolidation of senders’ deposits and the onward transmission of these funds to settlement
So how is it all achieved? As ever the back-office mechanisms
partners, and the deconsolidation of credits received from
which support these apparently straightforward front-office
settlement partners and their distribution to assigned recipi-
results are extremely complex, so much so that they are best
ents at a variety of overseas destinations.
considered in a series of stages, all of which have to be fitted smoothly together for the whole operation to produce the
Since all remittance agencies, be they formally constituted
desired result.
banks or Hawaladars operating in the informal sector, face the exploring which of the two is in a position to implement this
The collection of sterling in U.K. and the delivery of rupees in Mirpur
complex set of tasks with the highest level of effectiveness,
In front-office terms, Hawaladars engaged in collecting and
same set of technical and logistical challenges, it is worth
efficiency, and reliability.
dispatching remittance transfers necessarily work in close collaboration with a corresponding partner operating at the
A concrete example: the transfer of remittances from the U.K. to Pakistan’s Mirpur District
destination(s) to which those remittances are dispatched.
More than a third of a million people whose ancestral origins
the moment, implementing the component of the deal which
lie in Mirpur District are resident in the in the U.K., most of
matters to the customer, namely the delivery of rupees to the
whom keep in close touch with their kinsfolk back home. Not
recipient, is simply dependent on the swift and accurate trans
only are visits back and forth extremely frequent (PIA oper-
mission of the appropriate delivery instructions. However to
ates no less than ten flights into Manchester every week), but
send separate messages for every single transaction would be
many of the pioneer settlers who arrived in the U.K. forty
highly inefficient, consolidation is the order of the day. Hence
years ago have now retired, and are now taking the opportu-
each consolidating Hawaladar (and for clarity’s sake let us identify him as Ha operating in Birmingham9) will make up a
nity to use their accumulated savings to build splendid new houses for themselves back in Mirpur.8 As a result, remit-
list, setting out all daily transactions with over-the-counter
tances from U.K. to Mirpur are currently running at some-
customers and those passed on to him by a network of agents
where in the region of £250 million (approximately U.S.$ 450
and sub-agents operating in smaller Pakistani communities
million) per annum, the greater part of which passes through
elsewhere in the U.K. Then it is a simple enough task to fax the spreadsheet to his corresponding partner Hb, operating in
Hawala networks. Delivery is routinely achieved within 48
148 - The
Moreover if we leave the issue of settlement to one side for
Journal of financial transformation
Delivering migrant workers’ remittances
mentation of formal bureaucratic procedures of paper trails
has lead to an upsurge in intra-regional trade in manufactured
and audit inspections, that IVTS networks have characteristi-
goods. Finally, radical improvements in communications tech-
cally achieved a condition of self-regulation. Whilst reliance on
nology — including fax, internet, cell- and satellite phones —
self-regulation in this sense may seem as old-fashioned as it is
have made such transfers more efficient. Taken together,
insecure, it is worth remembering that so long as the networks
these developments rapidly gave Hawala operations a new
within which these transactions are implemented have an all
lease of life.
encompassing impact on their members’ lives, sanctions tions, can be as far-reaching as they are severe. Breach of
How do contemporary IVTS systems actually work?
against malfeasance, no matter how informal their foundatrust characteristically leads to swift and comprehensive
Although Hawala networks and other similarly structured
social ex-communication not just of the offender, but also of
IVTS operations are often described as providing under-
his entire family; and because these networks are now becom-
ground banking services to their clients, it is worth remember-
ing ever more transnational in character, flight offers fewer
ing that Hawaladars do not provide retail banking facilities,
and fewer opportunities for escape.
such as current accounts, taking deposits, or making loans. Instead IVTS networks are best understood as specialist forex
The gradual eclipse and subsequent revival of Hawala
agencies whose sole purpose is to facilitate value transfers over long distances and between different currency regimes,
Financial operations only survive if they can sustain a position
and in so doing specialize in the delivery of relatively small
of competitive advantage, and IVTS systems are no exception.
sums to recipients based in remote destinations in the devel-
When European adventurers initially broke into Indian Ocean trading systems in the 16th century they soon found that they
oping world.
had little alternative but to make extensive use of locally con-
Fulfilling that specialist role is far from straightforward. Even
stituted financial services, and for straightforward commercial reasons they continued to do so until the beginning of the 19th
though the gross annual flow of credits into any given labor
Century. But as Imperial domination became steadily more
lars per annum, closer inspection soon reveals that organizing
exporting country may well amount to several billions of dol-
comprehensive, European banks gained a dominant position
the swift, reliable, and efficient delivery of innumerable small
in major trading centers all around the Indian Ocean. As a
packets of value on a global scale is a major logistical chal-
result more indigenous providers of financial services were
lenge. With that in mind it is worth running through just what
driven steadily outwards towards the social, geographical, and
kind of activities the task necessarily entails.
economic periphery of the system. It goes without saying that in meeting this challenge However, in the post-colonial world four key developments
Hawaladars rarely become involved in the physical transfer of
have facilitated the resurgence of IVTS operations in general,
currency notes across international borders. Instead they use
and of Hawala in particular. Firstly, the imposition of over-tight
familiar banking processes of consolidation, deconsolidation,
exchange controls in many parts of the Indian Ocean region
and settlement whenever and wherever they possibly can.
has resulted in the emergence of a vibrant black market in
Hence Hawaladars’ collection, transmission, and delivery
foreign exchange. Secondly, as the number of labor-migrants
operations have three core elements, all of which are closely
from all over the region to the Gulf and Euro-America grew so
inter-connected.
did the amount of migrant remittances. Thirdly, the rapid growth of manufacturing activity in East and South East Asia
The first is information processing, which deals with provision
8 A detailed account of the history of mass migration from Mirpur District, as well as of the impact of the inflow of remittances on the local economy can be found in Ballard, R. “A case of capital-rich under-development: The paradoxical consequences of successful transnational entrepreneurship from Mirpur” in Gardner and Osella (eds.) Migration, modernity and social transformation in South Asia, New Delhi: Sage, 2004.
9 I have borrowed the notation for the identification of Hawaladars as Ha, Hb, etc. from the model set out in El Qorchi, Maimbo and Wilson “Informal Funds Transfer Systems: and analysis of the informal Hawala System” Washington DC: IMF/World Bank, 2003. However as my subsequent analysis shows, Hawala deals very rarely take the form of the simple bilateral exchanges which they set forth in their model.
149
Delivering migrant workers’ remittances
In these circumstances the outflow of migrant workers to the Gulf, the U.K, and the U.S, together with migrant workers’
just how the whole settlement process operates. Let us suppose that Ha in Birmingham has taken orders for the delivery
ting their remittances back home, has given a powerful boost
of Rs. 10 million in Mirpur, for which he has received £75,000 from his U.K.-based customers. Meanwhile Hc in Karachi has
to Karachi’s informal money market.
a customer who wished to purchase $100,000 in order to
As compared with Hawaladars serving the needs of migrant
has imported from China, whose manufacturer is expecting payment to be made into his U.S.$ account in Hong Kong. Ha
constant search for cheaper and more effective means of get-
settle the invoice for a consignment of televisions which he workers, their counterparts in Karachi operate in a very different kind of financial context. Not only are the majority of their customers drawn from the city’s business and professional elite, but such orders for U.S.$ are invariably on a substantial scale. Hence a single order for U.S.$ in Karachi may well generate a sufficient quantum of rupees to meet a week’s worth of daily out-payments in penny packages handles by Hb in Mirpur. Final settlement can only be achieved when the hardcurrency credits held by Ha in Birmingham have been transferred to the destinations specified by the customers of Hc in
and Hc separately approach a specialist settlement broker in Dubai, Hd, who calculates (for the sake of argument) that Rs. 10 million = U.S.$ 100,000 = £75,000, so generating an ideal opportunity to arrange a back-to-back swap. Hence Hd sets up a Hawala settlement in which Ha buys U.S.$ 100,000 on London money market through his bank which he promptly sends by TT to Hd’s account with Bank of America in New York; meanwhile Hc takes delivery of Rs. 10 million in cash
Karachi.
from his television-importing customer, which he promptly dispatches by road to Hb in Mirpur, thereby recompensing Hb
Dubai as a settlement hub
for the disbursements made in response to Ha’s previously faxed instructions; and last but not least Hd transfers
Most such deals are currently brokered in Dubai. Why Dubai?
U.S.$100,000 by TT from his account in New York to the tele-
Not only has Dubai taken advantage of the boom in the price
vision manufacturer’s account in Hong Kong, thus fully clos-
of oil to establish itself as the Gulf’s most important commer-
ing the circle.
cial entrepôt, but as a result of its close proximity to Bombay and Karachi, the stability of the Dihrams link to the dollar, and
However it goes without saying that this is only a hugely sim-
its consequent condition of unrestricted access to the global
plified example, and that in the real deals such settlements are
financial system, it has become South Asia’s premier offshore
a great deal more complex. Hence in the instance cited, there is every prospect that Hc in Karachi would approach Ha in
banking center. So it is that Dubai’s Exchange houses have come to play a key role as facilitators of remittance and com-
Birmingham directly, and having found a neat match between
mercially-driven Hawala settlements not just to serve the
their immediate financial requirements would do a simple
financial needs of the very substantial migrant-worker population in the oil-rich regions which surround it, but also on an
swap between themselves, such that there would be no need to look to Hd, who would inevitably take a small percentage
ever more global scale. Hence with their newly established
cut on the deal to facilitate such a settlement.
headquarters in Dubai, Hawala networks, which have their In practice, opportunities to implement neatly managed swaps
roots in mediaeval practices, have once again begun to oil the wheels of trade in the entire Indian Ocean region.10
involving such large sums are relatively rare, given that
A worked example
from which they seek to make deals in many different curren-
With all this in mind, a worked example, albeit much simpli-
cies in tranches which very rarely match up anything like so
fied, may provide a clearer and more concrete indication of
neatly as they do in the example cited. It follows that since
Hawaladars operate in such a wide variety of national arenas,
150 - The
Journal of financial transformation
10 Ballard, R. ‘Hawala Transformed: Remittance-driven Transnational Networks in the post-Imperial economic order’ in Maimbo and Ratha eds. Remittances: Development Impact and Future Prospects. World Bank Publications, Washington D.C. (2004)
Delivering migrant workers’ remittances
Mirpur, who can promptly set about organizing the local deliv-
reserves.
ery of rupees to recipients as instructed, and where necessary using his own network of agents and sub-agents to make deliveries to recipients resident in villages in the more remote parts of the District. Please note that there has been no transfer of value from Ha to Hb. Corresponding Hawaladars are ready and willing to take on such a position of indebtedness given the relationship of trust between them, but at some stage that condition must of necessity be discharged.
Organizing settlement Because the majority of migrant workers are drawn from relatively remote rural areas the most distributing Hawaladars operate in regions where there is no significant local demand for foreign exchange. That is certainly the case in Mirpur, so Hb rarely if ever has local customers interested in exchanging rupees for the sterling credits which he is constantly accumulating with Ha in Birmingham. However, in Pakistan’s commercial capital, a thousand miles away in Karachi there is an almost insatiable demand for access to hard currency, for which members of the local business elite are frequently prepared to pay premium rate. Hence the informal money markets in Mirpur and Karachi are effectively mirror-images of one another. This provides an obvious opportunity to organize and implement complementary back-to-back settlements, with the added risk of having to move the cash between these two cities, which in most cases require hiring of armed guards.
Karachi’s Hawala market Like many other developing countries, Pakistan has long suffered from a severe deficit in its balance of trade, which is only partially remedied by remittances and other invisible exports. Since Karachi is Pakistan’s commercial capital as well as the country’s principal entrepôt, a major manufacturing sector, the city has long supported a pent-up demand for foreign exchange to settle invoices for manufactured goods imported from overseas, to purchase education, medical treatment, and property overseas, to aid the wealthy to export capital to hardcurrency tax-havens beyond the purview of the Pakistani authorities, and to replenish the State Bank’s foreign-currency
151
Delivering migrant workers’ remittances
Regulation: Is Hawala broke? And does it need fixing?
authorities have undergone a sea-change since 9/11. Hence
Whilst there can be no doubt whatsoever that Hawala net-
nent of the regulatory agenda, measures to curb money-
works, as well as their Chinese and Latin American equiva-
laundering in general, and the financing of terrorist activities
lents, provide an excellent deal from the perspective of the
in particular, have suddenly leapt to the head of many agen-
many millions of migrant workers who regularly utilize their
cies’ concerns. Given that IVTS networks stand beyond the
facilities, the prospect that these informal, and currently
reach of any kind external regulation, alarmist accounts of
whilst customer-protection still remains a significant compo-
almost completely unregulated, systems might also have sig-
their vulnerability to criminal penetration are now regularly
nificant operational down-sides also needs careful consider-
articulated in all manner of official reports, most particularly
ation. It is, therefore, important to determine how secure such
those emanating from North America.
value transmissions are, as far as its retail customers are concerned, and how vulnerable such networks are to criminal
Yet despite these regularly articulated concerns, which have
penetration?
in turn led to the introduction of a swathe of anti-money laun-
Efforts to answer these questions have recently begun to pre-
tor, formal as well as informal, is now required to conform,
dering initiatives with which the entire financial services seccipitate a debate about how, on what basis, and to what ends
much less attention has yet been paid to determining just how
this increasingly significant corner of the global foreign
far the informal sector has actually been used for such nefar-
exchange market might best be regulated. Four equally press-
ious purposes. In a similar vein it is also worth noting that
ing questions have now begun to be posed within that debate:
there have been even fewer efforts to explore whether the raft of anti-money laundering initiatives, with which all institu-
■ To what extent do current Hawala networks have any in-
built guarantees of system security? ■ If these guarantees are either absent or inadequate, what
tions in the formally constituted financial services sector have to comply, have had any significant impact on their nominal targets. If, as I strongly suspect, their impact so far has been
measures might realistically be introduced to improve
more symbolic than real, it is high time to consider what alter-
current levels of system security?
native initiatives might be expected to produce the desired
■ Is there any way in which adequate and effective regulato-
result.
ry measures can be introduced whilst still retaining the strengths of such systems’ informal procedures? ■ Should all efforts be concentrated on driving all these
Before proceeding further it is also worth reminding ourselves that money-laundering is a catch-all term covering a wide
dangerously informal financial transfers into the properly
variety of different forms of financial malfeasance. Not only
regulated forex channels maintained by the formal bank-
does the extent and character of their ‘criminality’ vary enor-
ing sector?
mously, but there are excellent reasons for suggesting that only a small minority of the transactions so labeled are associ-
152 - The
How vulnerable are contemporary Hawala systems to criminal penetration?
ated either with drug-smuggling or with terrorist finances.
Yet, even if we accept that these informal systems have been
both individuals and corporations regularly devise as a means
Instead the vast majority are the outcome of strategies which
providing an efficient, effective, and above all a reliable
of confounding the efforts of national governments to tax
response to the logistical challenges thrown up by the excep-
their income, to impose excise duties on goods they wish to
tionally rapid growth in the North to South remittance deliv-
import or export, and/or to impose restrictions on their ability
ery market in recent years, the priorities of the regulatory
to export funds into alternative financial jurisdictions. The
Journal of financial transformation
Delivering migrant workers’ remittances
mix’n’match settlements are extremely complex to set up,
approach to the data which they choose to transmit between
they regularly require the lubricating assistance of brokers in
themselves, Hawaladars have typically taken immediate
Dubai if such exercises are to be swiftly and smoothly imple-
advantage of developments in communications technology.
mented. Moreover all the evidence of the resultant settlement
Hence in addition to landlines and mobile phones, Hawaladars
transactions which I have so far been able to obtain suggests
have moved rapidly onwards through fax, email, network por-
that they are conducted on a very substantial scale, since the
tals and secure Virtual Private Networks as a means of infor-
basic unit of account in making such deals appears to be
mation transmission amongst themselves, whilst all of the
U.S.$100,000.
larger operators, and an increasing number of smaller ones, have now begun to plug directly into inter-bank electronic
How secure is IVTS/Hawala as far as its retail customers are concerned?
value transfer systems, such as BACS and SWIFT.
Whilst billions of dollars are processed through Dubai-based
Given the manifest commercial efficiency with which contem-
IVTS/Hawala networks every year, it is striking that despite
porary Hawala systems cope with the logistical challenges
the absence of any kind of central registry, let alone a central
thorn up by migrant workers’ financial services requirements,
regulator, serious customer complaints are virtually unknown.
as well as the ever-greater technical sophistication of the
How can this be? The basic answer is quite straightforward:
transfer and settlement processes which they have developed
like the pre-colonial structures from which they evolved, con-
in order to meet those requirements, the use of the term
temporary Hawala networks are best understood as distrib-
‘informal’ as means of characterizing the system as a whole
uted systems in which system security is rooted in the trans-
needs to be approached with considerable care. In the first
nationally extended relationships of absolute trust which
place, any suggestion that Hawala’s informal character means
participating Hawaladars routinely maintain between them-
that it is ‘a system without records,’ as those who suspect that
selves. Given that all members of the network are active par-
the whole exercise is merely a convenient cover for criminal
ticipants in the whole enterprise, system security is not only a
activity frequently argue, falls by the wayside. Despite their
matter of collective concern, but is also articulated from the
commitment to reducing clerical procedures to a minimum,
bottom up, in sharp contrast to the top-down patterns of
there is no way in which Hawaladars could deal accurately and
external scrutiny and regulation which contemporary Western-
efficiently with transactions of the complexity to which remit-
style banking systems routinely deploy for that purpose.
tance transfer processes give rise without extensive recordkeeping. Secondly, and consequently, ‘informality’ is not to be
It is also worth emphasizing that such informal methods of
confused with back of the envelope procedures. On the con-
guaranteeing system security are far from being an outmod-
trary Hawaladars have been much quicker to make use of the
ed hangover from by-gone age. Quite the contrary, such
efficiency gains which can be extracted from recent develop-
practices appear to be the prime source of the system’s com-
ments in electronic communications than their rivals in the
petitive advantage, above all because they allow large
formal banking sector.
swathes of operationally redundant clerical and bureaucratic activity to be stripped out of the system. By doing so,
This immediately prompts a further question. Could it be that
Hawaladars are able cut down information transmission to
the competitive advantage currently enjoyed by contempo-
the minimum level required to implement the task in hand, so
rary IVTS networks is similar to the advantages Internet-based
enabling their core business of value transmission to be
retail banks have over their branch-based competitors that
implemented with the maximum level of speed, reliability, and
are struggling with the crippling costs of maintaining a pres-
efficiency. Moreover having adopted such a lean and mean
ence on the high street?
11 Robinson, J. The Sink: Terror, Crime and Dirty Money in the Offshore World Constable. London (2003) 12 Public Affairs (2004) The 9/11 Commission Report New York: Public Affairs.
153
Delivering migrant workers’ remittances
serious financial arsonists long since perfected the art of
bureaucratic procedures of questionable utility, most espe-
building smokeless fires. In these circumstances there are no
cially if those procedures stood comprehensively at odds with
short cuts. If the security services really want to track down
the way in which his business routinely operated. By the same
terrorists, and if law-enforcement officers really want to track
token it also makes little sense to demand conformity to a
down drug smugglers, neither cracking down on Hawaladars nor ever more elaborate KYC requirements are likely to be of
regulatory regime which was designed for use in a business which is wholly different in character13. Nevertheless, it is to
much use. If the authorities really want to catch terrorists and
everyone’s advantage, except for committed criminals, that
drug smugglers, rather than to generate headlines which sug-
Hawala and other similar IVTS networks should be brought in
gest to the public at large that something is being done, there
out of the cold. But that will only occur when much greater
appears to be no alternative to the hard slog of detective work
cognizance is taken of the way in which contemporary remit-
in which suspect transactions, no matter how complex and
tance-driven value-transfer systems actually operate, such
obscure, are tracked right through from one end to the other,
that appropriately-grounded and genuinely reality-checked
and arrests are made. The current scatter-gun approach,
regulatory procedures can be introduced.
which appears to be as much directed at the formal as the informal sector, may well be a vehicle for bringing successful
Meanwhile on quite another level altogether, there can be no
prosecutions against those who have failed to conform to post
doubt whatsoever that IVTS networks are highly effective
9/11 regulatory requirements. If the ‘war on terror’ goes no
service-providers in an increasingly significant corner of the
further than that, such surrogate successes appear to be little
global financial services sector. Moreover it should by now be
more than diversionary exercises in public relations. In such
quite clear that informal in this context should not be viewed
circumstances terrorists and drug smugglers will continue to
as either casual or unsophisticated. IVTS networks are in
wend their way through the global financial system with no
essence a highly efficient technical response to some very
greater difficulties than they have so far enjoyed.
substantial logistical challenges. As such they have emerged as very significant competitors to more mainline institutions
Conclusion
which operate on a more formal basis. But as Gerry Adams
In my experience the vast majority of Hawaladars have no
said of Irish republicanism, ‘it’s not going to go away, you
great objection, at least in principle, to subjecting themselves
know.’ Just as Japan, and more recently China, have reversed
and their operations to regulatory scrutiny, especially if the
the terms of trade between East and West through their inno-
central aim of those regulations is to ensure that their facili-
vations in the manufacturing sector, we should also not be
ties are not being surreptitiously utilized by terrorists and
surprised if similar challenges from below also begin to
drug smugglers. After all, like all other operators in the finan-
emerge in the financial sector. This is not to suggest that there
cial services sector they have businesses to run, and client
is any likelihood that Hawaladars will be turning the way in
confidence to maintain. Blithely processing criminal transac-
which the entire established financial services sector operates
tions would in no way be worthwhile in straight commercial
on its head. But, it is to suggest that in the specific field of
terms.
long-distance cross currency value transfers, the formal sector has a great deal to learn from its more informal competi-
However, given that they have businesses to run, and that position of competitive advantage arises precisely from their capacity to implement long-distance value-transfers on a comprehensively lean-and-mean basis, no Hawaladar would ever wish to be saddled with a requirement to implement elaborate
154 - The
Journal of financial transformation
tors.
Delivering migrant workers’ remittances
resultant contradictions are of long standing. Just as states
The fact that those funds were transferred through the formal
have always needed to tax their subjects in order to finance
banking system also raises another important wider question:
their activities, their subjects have done their best to avoid
to what extent are transfers made through informal sectors,
them. Consequently, if Hawala networks also provide support
as is widely assumed, ‘more anonymous’ than those sent
for such evasive strategies, which they almost certainly do for
through more established banking channels? In fact there is
people of the developing world, we should not be greatly sur-
good reason to suppose that the opposite is true. By defini-
prised. However such strategies are in no sense unique to the
tion, Hawaladars normally know their customers on a per-
informal sector. On the contrary, it seems quite clear that all
sonal basis. When such customers transfer substantial sums
the biggest players, as well as the vast bulk of the funds laun-
of money, the Hawaladars will normally have a pretty shrewd
dered in this sense, slip with ease between on- and offshore
idea of the sources of those funds, and the purposes for which
institutions which are firmly located within the formally constituted financial sector.11
they have been dispatched. Moreover, if he knows that there
AML/CFT
aboard will be well aware that by implementing that deal he is
What, though, about transactions at the much more serious
implicitly confirming the bona fides of the whole transaction
end of the money-laundering spectrum, drug smuggling and
as far as all his partners in that Hawala network are con-
is a prospect that those funds have been acquired on a seriously illicit basis then the Hawaladar who takes those funds
terrorist financing? Here we move into even shadier financial
cerned, needlessly putting the integrity of the system as a
arenas. By contrast with the vast sums of money regularly
whole at risk. In other words, not only are Hawaladars likely to
transferred around the globe to facilitate tax evasion in all its
be aware of which of their customers might be engaged in
various forms, the volume of cross-currency transfers gener-
shady deals, but they are also under a significant degree of
ated by drug-smugglers seems to be relatively small. For
systematic pressure to keep clear of such transactions.
example, although Afghanistan may currently be the source of around 80% of the world’s heroin supplies, it is quite
By contrast there are good reasons for supposing that serious
unreasonable to suppose that anything more than a tiny pro-
criminals have a much better opportunity of preserving their
portion of the profits accruing from its sale in Europe and
anonymity by dealing with the formal banking sector. Post 9/11
North America ever come anywhere near the country in
Know-Your-Customer AML procedures with which all major
which the poppy crop is grown. Drug smugglers may indeed
institutions in the financial services industry are now required
need to launder their ill-gotten gains, but it makes no sense
to conform are expensive to implement, and customers may
for them to seek to do so by investing in the economies of
often be annoyed by the requirement to produce a variety of
failed and near-failing states. Meanwhile there is a growing
documents to verify their personal identity. However there is
awareness that the sums required to finance terrorist atroci-
no reason to suppose that those determined to slip beneath
ties are so small that many commentators have likened the
the net would have any difficulty whatsoever in doing so.
task of identifying such transfers to looking for a needle in a haystack. The FBI estimates that the cost of mounting the
Money, it is worth remembering, is merely a medium of
September 11 attacks was somewhere between U.S.$175,000
exchange. It is neither black nor white, and neither guilty nor
and U.S.$200,000, very small beer in the context of the
innocent. If the authorities are really concerned about track-
global forex market. Moreover the bulk of these funds were
ing down serious criminal activities it is becoming increasingly
dispatched to the U.S. by wire transfer into an account which
clear that to do so solely on the basis of seeking to identify
two of the perpetrators opened at the SunTrust Bank in Venice, Florida.12
‘suspicious transactions’ is highly unlikely to provide positive
13 In the Netherlands and Germany, for example, current regulations treat value transmission agencies as banks, and hence require them to maintain capital reserves and to use audit procedures of the kind which are entirely appropriate for institutions which offer a comprehensive banking services. In the face of such requirements hawaladars have no alternative but to conduct their operations ‘underground’.
results. Whilst smoke does not necessarily conceal a fire, most
155
Future
Check 21 and the migration to electronic payments
Adam Dener Partner, Capco
Abstract This article explores the increasing changes in payment and payment flows driven by regulation, customer behavior, and technology. The implementation of Check 21 legislation in the fourth quarter of 2004 will further accelerate the migration from paper to electronic payments. This change is expensive and will impact banks. Banking institutions will need to invest in the front-, middle-, and back-office infrastructures with total industry costs to exceed U.S.$10 billion over the course of the decade. The increasing migration also raises risks to the banks as businesses are better able to gain benefit from changes and in turn will be better able to collect funds. The acceleration of collection has significant potential negative consequences on bank revenues from deposits and payment products.
157
Check 21 and the migration to electronic payments
permit and/or likely enable stopping the flow of original
increased expenses to support redefined processes, incorpo-
checks in the collection and/or return process, reducing the
ration of imaging and digital conversion into the processes,
physical processing and shipping of checks, establishing
and incorporation of imaging technologies. These legacy costs
agreement to collect or return checks electronically, and cre-
will increase cost-per-check processing in the short term.
ating network agreements from thousands of banks to accept.
Initial transportation cost reductions will not adequately offset the increased costs, while redundancy costs inherent in
The detailed provisions of The Check 21 Act incorporate a
today’s processing environments will significantly increase.
framework that includes the definition of substitute checks, the establishment of a new negotiable instrument, and an
We estimate that the industry’s near term investment costs
agreement that a substitute check is a paper reproduction of
will be approximately U.S.$10 billion over the next 3 to 5 years.
the original check and can be processed as such. The paper
This figure includes new imaging input hardware, storage,
reproduction must, however, meet industry standards for sub-
retrieval, and software supporting input, storage, and retriev-
stitute checks, include an image of the front and back of the
al. In addition, it includes network connectivity supporting
original check, and be MICR encoded and machine readable.
input, storage, retrieval, and presentation as well as the redundancy costs of current check processing environments, includ-
A substitute check is the legal equivalent of the original provided that it accurately represents all the information on the front and back of the original check, bears a legend stating it is a legal copy of your check and that you can use it the same way you would use the original, and that no agreement is needed to transfer a legally equivalent substitute. Check 21 does not, however, specifically provide for electronic exchange bank to bank although implicit in the legislation is the ability for this type of service provision.
Incorporating imaging into check processing — A U.S.$10 billion problem Today’s check processing environment involves a complex network of processes involving significant labor, expensive equipment and systems, transportation, and inter-party affiliations. A key element of determining the necessary check processes is the predicated location of presentation of the check by or to a bank, which as discussed previously, is a legal negotiable document. Check 21 complicates this situation by expanding the scope of current and potential processes while increasing the need for technology support for the equipment and systems that support them. The short-term impact of Check 21 compliance will be
158 - The
Journal of financial transformation
ing labor.
Check 21 and the migration to electronic payments
Recently approved Check 21 legislation will be implemented in
tem without mandating receipt of checks in electronic form,
October 2004, the objective of which is to reduce systemic
and to improve the overall efficiency of the Nation’s pay-
risk associated with physical check movement and to provide
ments system, and for other purposes.’ Public Law 108–100, 108th Congress.
a mechanism to reduce the costs associated with check payment processing. Check 21 creates a newly acceptable payment instrument and provides a foundation for the imaging of
The Check Truncation Act was signed into law on October 28,
checks as the basis for creating new payment instruments and
2003 by President Bush. The law, commonly referred to as
their acceptance by other banks. While the Act does not spe-
Check 21, provides a legal framework for banks to accept sub-
cifically require that images be presented bank to bank, the
stitute forms of checks effective October 2004. The likely
legal and operational foundation exists for this evolution.
initial outcomes of the legislation are the reduction of systemic risk of lost checks due to business continuity failure,
However, closer scrutiny of the legislation point to significant
such as 9/11, and cost reduction by allowing banks to agree to
business risks from both a cost and revenue perspective.
electronic presentation, thereby eliminating the need for
These include:
physical presentation and transportation of checks for settle-
■ Industry costs — Check 21 will, over the next 3 to 5 years,
ment.
cost the industry U.S.$10 Billion to manage banking industry compliance implementation and redundancy costs
It is the broader implications of the law, however, that create
associated with reductions in legacy investments in check
even more significant benefits as well as risks for banks. The
processing.
legislation will likely serve as an enabler that will form the
■ Cost savings — Post-implementation will allow for pro-
cessing cost savings and, with the embrace of digital
basis of check digitization leading to digital presentation bank to bank.
presentation, significant transportation cost savings. ■ Float and fee revenue loss — Check 21 poses significant
revenue risk by migration of check processing to ACH or
Network and servers
digital checks at significantly lower fees, changes, and reductions in check float. ■ Insufficient funds fee loss — The legislation also has the
ability to create funds verification services which could reduce insufficient funds processing and fees (NSF).
Depository bank
Paying bank Presentation
Check processing
Systems & I/F
Systems & I/F
Check processing Process
To date, most observed industry efforts have been compliance, as opposed to business, focused. Given the scope of
Paper checks
Clients/ customer service
Consumer
Merchant
both the cost and revenue impact, I would recommend that
Clients/ customer service
bank Check 21 programs be aggressively managed as business programs. There is significant opportunity, as well as risk, based on which approach is ultimately pursued.
Image archive
What is Check 21? ‘An Act to facilitate check truncation by authorizing substitute checks, to foster innovation in the check collection sys-
Data flow Image flow Point of impact / enhancement Check 21 creates a legal and operational foundation that will
Figure 1: Processes where Check 21’s impact will need to be addressed
159
Check 21 and the migration to electronic payments
and disbursing float — they also lower transaction revenues
tation and with bank-to-bank presentment. As noted in the survey, most bank imaging efforts have been focused on inter-
and margins.
nal processing. While certain industry efforts have been creCorporations have been willing to accept the positive benefits
ated that are focused on inter-bank capabilities, such as
of converting check-based receipts to ACH transactions,
archiving, the advent of bank-to-bank presentment capabili-
generically referred to as Check Conversion. This is also borne
ties and image capture at point of presentation will have significant impact on check-processing economics.
Image capabilities of image enabled banks Statements Rejects
Returns
Posting
The bulk of retail check presentation occurs at bank branches,
National banks
68%
49%
26%
34%
retail merchants, and check-processing merchants. Such pre-
Regional banks
47%
34%
27%
22%
sentation is physical across multiple channels (e.g. ATM, teller, POS) and represents a significant opportunity to transfer pro-
Who would you use as a processor? 3rd Party FRB Consortium RPA National banks 27% 4% 0% 4% Regional banks
45%
35%
32%
25%
cessing costs to the presenter by incorporating imaging into Another bank 12% 11%
presentation processing. This capability will be of benefit to bank branches and retailers with significant collection requirements. It is likely that some retailers, including banks, may drive bank providers to enable capabilities for their benefit — to expedite their collection and float concentration capabilities —
out by the rapid growth of check conversion transactions in
to the detriment of less sophisticated disbursers and disburs-
the last 24 months. To date, check conversion has had little
ing banks. The incorporation of imaging technology at point of
impact on corporate payments, since it has, by regulation,
presentation will both enable significant savings and opportu-
been limited to consumer initiated transactions. This situation
nity as well as change behaviors of end-user customers.
is beginning to change, as the U.S. Treasury has recently issued proposed regulations that will allow it to convert com-
In addition, bank-to-bank presentation (B2BP) for collection
mercial payments to ACH transactions as well if it wishes to do
will drive further cost efficiencies. By incorporating imaging
so. The implementation of Check 21 truncation, which has no
POPIC with electronic presentation processing, settlement
similar limitation, will only likely accelerate the movement
can be enabled by the presenter automatically without direct
towards electronic payments.
staffing (or transportation) other than support for the custodial function of imaged checks.
The result of these customer and regulatory transitions is decreasing margin exacerbation in check processing, as check
In addition to business continuity related risk reduction, Check
volume declines. Bank attempts to maintain margins by rais-
21 implementation will offer significant long-term financial
ing check prices could drive more transactions to electronic
incentives for the industry, including reduced transportation
media or competitors, compounding the problem. Such events
costs, process staffing, and labor requirements.
could further strain the significant investments in check infrastructures and economics.
160 - The
Analogue/paper to digital payments: The risk of lost float and revenue reductions
Point of presentation image capture (POPIC)
Check 21 regulations enter the business cycle at a precarious
Another likely change related to Check 21 is lower revenues
time for the banking industry. Already under economic pres-
predicated by the retail needs of customers at point of presen-
sure exacerbated by the recent recession and credit cycle, the
Journal of financial transformation
Check 21 and the migration to electronic payments
The banking industry has the ability to control elements of
The bulk of imaging integration is now focused on statement
this impact through staging, predicated on a redefinition of
processing, with approximately half as much automation in
processing. However, these impacts could also be driven by
reject, return, and posting processing. In addition, respon-
other banks and key customers who could hasten and even
dents indicated a bias against support from other providers
define implementation staging by pursuing legislatively sup-
for imaging enablement through incorporation of service pro-
ported operational benefits.
visions (less than half of the banks responded in favor).
Figure 1 notes the processes where Check 21’s impact will need
Cost savings: The payoff from imaging?
to be addressed. These include at point of retail sale (cus-
The manifestations from bank compliance implementation
tomer to merchant), at point of branch presentation (ATM or
and the benefits offered by Check 21 have the potential to
teller), check processing input, check processing system, cus-
significantly impact bank costs, revenues, and even customer
tomer service support, bank to bank presentation, and image
behaviors.
archive and retrieval.
ACH adoption Our 2003 industry survey on preparedness for check process-
In commercial payments, the transition to ACH has accelerat-
ing and image automation discovered that only the largest
ed with conversion increasing significantly in the past 24
banks have begun to incorporate imaging technology into
months. The Federal Reserve has noted ACH growth recently
their services. Only 65% of overall respondents indicated
at more than 15%. While ACH has traditionally offered lower
some level of image enablement. When the respondents were
direct processing costs, commercial payers have been slow to
broken down into national and regional banks we found that
transfer due to the potentially negative impact on working
while 90% of the former group are image enabled, only 60%
capital, in the form of lost float and control over payment
of the members of the latter group are.
dates. Large, sophisticated companies have historically utilized internal and external resources to analyze the benefits of using checks versus ACH payments, and changed those payment flows they felt beneficial. Disbursing corporations have the potential to lose float benefits as collection cycles can be reduced through utilization of Check 21-aided digitization and electronic presentation. As a result, companies utilizing check float will likely modify their processes to neutralize the lost float benefit and control. ACH mechanisms can enable incorporation of disbursing lags to attempt cash flow neutrality. Alternatively, these companies may accept the lost float as an accomplished reality and migrate towards less expensive ACH payments. The impact of this conversion is significant for banks. While ACH services offer lower processing costs — this transition stranding significant investments in check processing infrastructures and shared benefits of collection
161
Check 21 and the migration to electronic payments
traditional commercial business is experiencing client conversion-based adoption of ACH while consumer business is under margin pressure due to increased competition for account ownership amid rising costs for new branches. Compounding these challenges, POPIC implementation and bank-to-bank presentation capabilities have the potential to significantly reduce bank revenues via reductions in float related to check collection and potentially altering processes related to insufficient funds (NSF) and transaction fees.
Float revenue reduction Disbursers and banks benefit from check float in three distinct cycles, mail float, the cycle where the check is in transit to collector (in transactions where presentation is not at POS); Processing float, the cycle where the check is being processed by the collector and by the collector’s bank; and clearing float, the cycle where the check is being moved from depository to paying bank for final settlement and funds release. Dependent upon the method of presentation, a significant number of checks that are presented via POPIC incorporating B2BP would potentially reduce the processing and clearing cycles. This cycle reduction would have significant impact in speeding up available collections, thereby reducing funds balances among disbursers and their banking providers. Federal Reserve statistics offer significant insight into the potential impact of Check 21 implementation. Monthly transaction deposit data (i.e. non-savings-accounts for commercial and retail customers) seem to be relatively flat over the past 13 years. Estimated float revenues generated from those deposits (i.e. Fed Funds rates) shows the net impact of the reducing interest rate environment. These impacts are significant but will be increased through further ACH and imaging based conversion. Check float has been decreasing as a result and further product transition and behavior change is likely to significantly reduce revenue potential. Analysis on current collection trends suggests continued downward pressure on the collection cycle, further
162 - The
Journal of financial transformation
pressuring float opportunity as well.
163
Guidelines for manuscript submissions Guidelines for authors
Manuscript guidelines
In order to aid our readership, we have established some guidelines to ensure that published papers meet the highest standards of thought leadership and practicality. The articles should, therefore, meet the following criteria:
All manuscript submissions must be in English.
1. Does this article make a significant contribution to this field of research? 2. Can the ideas presented in the article be applied to current business models? If not, is there a road map on how to get there. 3. Can your assertions be supported by empirical data? 4. Is my article purely abstract? If so, does it picture a world that can exist in the future? 5. Can your propositions be backed by a source of authority, preferably yours? 6. Would senior executives find this paper interesting?
Subjects of interest All articles must be relevant and interesting to senior executives of the leading financial services organizations. They should assist in strategy formulations. The topics that are of interest to our readership include: • • • • • • • • • • •
Impact of e-finance on financial markets & institutions Marketing & branding Organizational behavior & structure Competitive landscape Operational & strategic issues Capital acquisition & allocation Structural readjustment Innovation & new sources of liquidity Leadership Financial regulations Financial technology
Manuscript submissions should be sent to Shahin Shojai, Ph.D. The Editor Editor@capco.com Capco Clements House 14-18 Gresham Street London EC2V 7JE Tel: +44-20-7367 13 21 Fax: +44-20-7367 1001
164 - The
Journal of financial transformation
Manuscripts should not be longer than 5000 words each. The maximum number of A4 pages allowed is 10, including all footnotes, references, charts and tables. All manuscripts should be submitted by e-mail directly to the editor@ capco.com in the PC version of Microsoft Word. They should all use Times New Roman font, and font size 10. Where tables or graphs are used in the manuscript, the respective data should also be provided within a Microsoft excel spreadsheet format. The first page must provide the full name(s), title(s), organizational affiliation of the author(s), and contact details of the author(s). Contact details should include address, phone number, fax number, and e-mail address. Footnotes should be double-spaced and be kept to a minimum. They should be numbered consecutively throughout the text with superscript Arabic numerals. For monographs Jensen, M., Corporate Control and the Politics of Finance. Journal of Applied Corporate Finance (1991), pp. 13-33. For books Copeland, T., T. Koller, and J. Murrin. Valuation: Measuring and Managing the Value of Companies. John Wiley & Sons, New York, New York (1994). For contributions to collective works Ritter, J. R., 1997, Initial Public Offerings, in Logue, D. and J. Seward, eds., Warren Gorham & Lamont Handbook of Modern Finance, South-Western College Publishing, Ohio. For periodicals Griffiths, W., Judge, G., 1992, ‘Testing and estimating location vectors when the error covariance matrix is unknown,’ Journal of Econometrics 54, 121-138. For unpublished material Gillan, S., and L. Starks. Relationship Investing and Shareholder Activism by Institutional Investors. Working Paper, University of Texas (1995).
Request for papers — Deadline January 28th, 2005 The world of finance has undergone tremendous change in recent years. Physical barriers have come down and organizations are finding it harder to maintain competitive advantage within today’s truly global market place. This paradigm shift has forced managers to identify new ways to manage their operations and finances. The managers of tomorrow will, therefore, need completely different skill sets to succeed. It is in response to this growing need that Capco is pleased to publish the ‘journal of financial transformation.’ A journal dedicated to the advancement of leading thinking in the field of applied finance. The journal, which provides a unique linkage between scholarly research and business experience, aims to be the main source of thought leadership in this discipline for senior executives, management consultants, academics, researchers, and students. This objective can only be achieved through relentless pursuit of scholarly integrity and advancement. It is for this reason that we have invited some of the world’s most renowned experts from academia and business to join our editorial board. It is their responsibility to ensure that we succeed in establishing a truly independent forum for leading thinking in this new discipline. You can also contribute to the advancement of this field by submitting your thought leadership to the journal. We hope that you will join us on our journey of discovery and help shape the future of finance.
Shahin Shojai Editor@capco.com
For more info, see page 162
© 2004 The Capital Markets Company. VU: Shahin Shojai, Prins Boudewijnlaan 43, B-2650 Antwerp All rights reserved. All product names, company names and registered trademarks in this document remain the property of their respective owners.
165
Design, production, and coordination: Cypres — Daniel Brandt and Pieter Vereertbrugghen Š 2004 The Capital Markets Company, N.V. All rights reserved. This journal may not be duplicated in any way without the express written consent of the publisher except in the form of brief excerpts or quotations for review purposes. Making copies of this journal or any portion there of for any purpose other than your own is a violation of copyright law.
Four Key Questions for Wealth Managers
1 2
3 4
Rising costs, heavier regulation, tougher clients in a tougher market – how can a Private Banking business cope?
Can we respond quickly enough to all these changes?
SEI knows about the competitive benefits of outsourcing in private banking institutions.
Should we be redefining our private client value proposition?
To get a copy of SEI’s special report ‘Outsourcing and the European Wealth Management Market’*, call Francis Jackson on + 44 (0)207 297 6308, or Email fjackson@seic.com How do we decide what is best done in house, and what needs to be outsourced?
Hear what the industry is saying. * Research amongst CEOs, CFOs and Senior Executives of wealth management institutions in 10 European Countries SEI Investments (Europe) Ltd, 4th Floor, The Economist Building, 25 St James’s Street, London SW1A 1HA is regulated and authorised by the Financial Services Authority
SEI. New Ways. New Answers.
www.capco.com
Antwerp
T +32 3 740 10 00
Bangalore
T +91 80 527 0353
Boston
T +1 617 262 1135
Frankfurt
T +49 69 9760 9000
London
T +44 20 7367 1000
New York
T +1 212 284 8600
Paris
T +33 1 47 55 30 90
San Francisco
T +1 415 445 0968
Singapore
T +65 6395 6998