Financial stability report of Turkey by Business Turkey

Financial Stability Report May 2017 Volume 24

THE CENTRAL BANK OF THE REPUBLIC OF TURKEY

Head Office Anafartalar Mah., 7sƟklal Cad. 10 Ulus, 06050 Ankara, Türkiye

Tel: (+90 312) 507 50 00 Fax: (+90 312) 507 56 40 Telex: 44033 mrbrt tr; 44031 mbdõ tr World Wide Web Home Page: http://www.tcmb.gov.tr E-mail: bankacilik@tcmb.gov.tr, info@tcmb.gov.tr

ISSN 1306-1232 ISSN 1306-1240 (Online)

This report, which is aimed at informing the public, is based mainly on March 2017 data. Nevertheless, the Report includes developments and evaluations up to its date of publication in Turkish. The full version of this text is available on the CBRT website. The CBRT cannot be held accountable for any decisions made based on the information and data provided therein.

Foreword The coordinated policies implemented since the last volume of the Financial Stability Report have been effective in limiting macro-financial risks. In this period, downside risks to economic activity decreased considerably and the financial system maintained its strong outlook.

The Central Bank of the Republic of Turkey bolstered the resilience of our economy with its tight stance regarding the inflation outlook and supportive attitude towards financial stability.

It is my hope that the 24th volume of the Financial Stability Report, which presents a discussion of the global and domestic macroeconomic outlook as well as the most recent developments regarding financial stability, will be of benefit to all readers.

Murat Ă&#x2021;ETÄ°NKAYA Governor

Contents Overview .................................................................................................................. i I. Macroeconomic Outlook .................................................................................... 1 I.1 International Developments ...................................................................................... 2 Box I.1.I Protectionist Policies in International Trade and Possible Effects ......... 6 Box I.1.II Financial Instruments International Financial Reporting Standards 9: Expected Credit Loss ................................................................................. 9 I.2 Domestic Developments .......................................................................................... 13

II. Non-Financial Sector .......................................................................................... 17 II.1 Household Developments ....................................................................................... 18 Box II.1.I Gold as an Investment in the Financial System and Gold Banking in Turkey ............................................................................ 23 II.2 Real Sector Developments ..................................................................................... 26 Box II.2.I Rediscount Credit and Amendments Made to the CBRT Regulations on Rediscount ............................................................... 33

III. Financial Sector .................................................................................................. 36 III.1 Credit Developments and Credit Risk .................................................................. 37 III.2 Liquidity Risk ............................................................................................................... 45 Box III.2.I Net Stable Funding Ratio ......................................................................... 51 III.3 Interest Rate and Exchange Rate ....................................................................... 54 III.4 Profitability and Capital Adequacy ..................................................................... 56 Box III.4.I Exchange Rate Developments and The Capital Adequacy Ratio 59

IV. Special Topics .................................................................................................... 61 IV.1 Global Liquidity and Regional Distribution of Cross-Border Banks ................. 61 IV.2 Measures on Corporate Sectorâ&#x20AC;&#x2122;s Access to Finance ...................................... 71 IV.3 Drivers of Credit Dollarization ............................................................................... 77 IV.4 Effects of Retail Loan Regulations ........................................................................ 84 IV.5 The Role of Bank Characteristics in the Interest Rate Transmission................ 92

Appendix ............................................................................................................... 100 Charts, Tables and Figures .................................................................................. 101 Abbreviations ........................................................................................................ 105

Overview Since the November 2016 Financial Stability Report, global economic activity has gained momentum and an increasingly stable global financial environment has been observed. Domestically, downside risks related to economic activity have significantly decreased owing to the incentives and supportive measures taken. In the final quarter of 2016, the economic activity compensated for the losses compared to the previous period, and the recovery trend is expected to gain further pace in 2017 with accommodative macroprudential policies, fiscal policy and credit incentives to be implemented. These incentives contribute to financial stability by supporting the well-functioning of the credit channel. On the other hand, the CBRT has significantly tightened monetary policy starting from January to contain the upside risks on inflation. The firm stance of monetary policy resulted with a clear decline of currency volatility in the succeeding period of the policy response.

The uncertainty in the US financial markets, which were heightened in the aftermath of the presidential elections of the US, abated in the first quarter of the year. As a result of investorsâ&#x20AC;&#x2122; expectations for expansionary fiscal policy, increases in infrastructure investments and regulatory rollbacks in the financial system, equity prices increased substantially in the US. Currently, amid diminishing uncertainty about the Fed's policies, expectations continue that the tightening of the monetary policy will diffuse over an extended period. Global market players expect that the European Central Bank and the Bank of Japan will maintain low interest rates. The risk premiums have fallen and the volatility in the financial markets has decreased as a result of stronger global economic growth prospects and the measured decline in uncertainty about monetary policies. These factors also led to a surge in global risk appetite. The optimism for the global markets and stronger growth prospects in emerging market economies have had a positive impact on the risk perceptions pertaining to these markets. This positive impact increased portfolio flows to emerging markets in recent months, which in turn started compensating for the losses in the value of domestic currencies and other asset prices in the last quarter of 2016 (Charts 1 and 2). Chart 1

Chart 2

Weekly Capital Flows to Emerging Markets

Exchange Rate Indices

(Billion USD, 4-Week Cumulative)

108

161 USD Index

106 10

159

JP Morgan EM Currency Index (RHA)

157

104

155

102

153 100

-10

151 98

149

-20

04.17

02.17 03.17

01.17

12.16

11.16

10.16

09.16

08.16

07.16

143

06.16

145

05.16

05.17

03.17

01.17

11.16

09.16

07.16

05.16

03.16

01.16

11.15

09.15

07.15

05.15

03.15

01.15

-40

147

04.16

Equity Markets

02.16 03.16

Bond Markets

01.16

-30

Note: JP Morgan EM Currency Index is inverted.

Source: EPFR (Latest Data: 05.17)

Source: Bloomberg (Latest Data: 05.17)

Financial Stability Report - May 2017

The banking sector balance sheets have continued to expand thanks to the recent recovery in economic activity, incentives for access to finance and prospects of moderate growth. The share of loans remained high in bank balance sheets. Credit growth was supported by the macroprudential regulations introduced for consumer loans, the recent measures towards supporting the financing of the corporate sector and the arrangements made in the cost, collateral and maturity conditions of loans.

Corporate loans have

displayed a significant rise particularly after the second half of March 2017 as corporate loans backed by the loan guarantees provided by the Credit Guarantee Fund (KGF) accelerated (Chart 3). Thus, the recovery in the total credit volume was mainly driven by the TL corporate loans backed by the KGF. Housing loans and consumer loans have also contributed to the recovery in loans. It is expected that these developments will continue to make a positive contribution to the economic activity in the upcoming period. The fact that credit growth is mainly driven by the rise in corporate loans, mitigates the negative impact of the recovery in domestic demand on the current account deficit. Macroprudential policies, the tightening in the monetary policy stance and other measures and incentives introduced have been the factors that have affected loan prices (Chart 4). Chart 3

Chart 4

Annual Loan Growth

Loan Rates and Regulations

(Percent, FX-Adjusted)

(4-Week Moving Average, Percent)

Total Loans

Corporate

Retail

Regulation on Corporate Loans Regulation on Consumer Loans Corporate Loan Rates Consumer Loan Rates

20 20 15

10 5 0

04.17

01.17

10.16

07.16

04.16

01.16

10.15

07.15

04.15

01.15

10.14

07.14

04.14

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Note: FX loans are adjusted using a dollar-euro basket, and FXindexed loans are included in FX loans. Growth rates are calculated over monthly values until March 2017, and weekly values for April.

Note: Consumer and corporate loan rates refer to the rates of new loans.

Source: CBRT (Latest Data: 04.17)

As of the second half of 2016, owing to the fluctuations in foreign exchange rates stemming from local and global developments and increase in the costs of financing from abroad, corporate sectorsâ&#x20AC;&#x2122; foreign debts decreased with a tendency to switch from foreign currency (FX) loans to TL. The changes in the composition of loans coupled with the improvement in the net foreign exchange position of the corporate sector resulting from the measures taken by relevant institutions and the increase in the share of long-term foreign liabilities are considered as positive developments with respect to financial stability. The data monitoring system established at the CBRT is expected to help obtain healthy data on firmsâ&#x20AC;&#x2122; foreign exchange risk and bring about an improved data monitoring framework. In line with the objective of providing a structural underpinning to financial stability and price stability, it is important to reinforce cooperation and coordination among all participants in the

Financial Stability Report - May 2017

financial system to enhance the systemic risk measurement capacity, to ensure prudent borrowing and to contribute to risk management practices in the upcoming period.

The banking sector maintains its strong liquidity position and remains resilient to a possible liquidity risk. In light of the funding composition of the sector, a significant portion of TL-denominated resources is composed of core liabilities, which in turn contributes to the liquidity position. The share of non-core liabilities in total liabilities has remained stable in recent years (Chart 5). Despite a moderate decline in external borrowing due to domestic demand-driven factors in the recent period, maturities of non-core funding items have continued to increase as a result of the measures in practice, and hence strengthened the resilience of the banking sector against possible global liquidity shocks. As a reflection of the stable course of non-core liabilities, the loan to deposit ratio has remained flat in recent years (Chart 6). The core liabilities and improvement in banks' profitability, henceforth their internal capital generation have been key sources of credit growth.

Chart 5

Chart 6

Ratio of Non-Deposit Funding to Funding Sources

Loan/Deposit Ratio

(4 Week Moving Average, Percent)

Domestic Foreign Total (RHA)

(4 Week Moving Average, Percent)

140

Total

130 120

110 100 90 80

Source: CBRT (Latest Data: 03.17)

04.17

10.16

04.16

10.15

04.15

10.14

04.14

10.13

04.13

10.12

04.12

10.11

04.11

10.10

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Source: CBRT (Latest Data: 04.17)

The banking sector retains its healthy asset quality. The total nonperforming loan (NPL) ratio, one of the key indicators, has reached a plateau, whereas the credit restructuring has occurred mostly in performing loans in a way to include credit characteristics such as maturity, collateral and pricing, and the restructuring in NPLs has been modest. Recent data indicate that the growth rate of loans under close monitoring has begun to decline after a hike in the third quarter of 2016. Credit risks for these loans are anticipated to remain at a reasonable level in the upcoming period due to the expectations of a recovery in the economic activity as well as the incentives given to SMEs and the collaterals provided through the KGF (Chart 7 and 8). The current level of the banking sector capital and the recent recovery in profitability support banks' capital structure and their lending capacity.

iii

Financial Stability Report - May 2017

Chart 7

Chart 8

NPL Ratios

Corporate NPL Ratios

(Percent)

5.0

Total Corporate Retail

4.5

Total Large SME

4.5 4.0

4.0

3.5

3.5 3.0

3.0

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

09.14

1.5

06.14

2.0

12.13

2.0

03.14

2.5

Source: CBRT (Latest Data: 03.17)

The upward trend in banking sector profitability observed in 2016 has continued through 2017. The recovery in lending appetite, the improvement in net interest income and the persistent measures to minimize non-interest expenses have all contributed to the profitability of the sector (Charts 9 and 10).

Profitability is expected to continue in the

upcoming period as a result of the accelerated loan volume, increasing interest and commission income and the positive contribution of net interest margins. Chart 9

Chart 10

Return on Assets (ROA) and Return on Equities(ROE)

Spread between Loan and Deposit Rates

(Percent)

Dispersion of ROA (10 Largest Banks) 3

Return on Assets (Sector) Return on Equities (Sector) (RHA)

Spread (RHA) TL Corporate Loans TL Deposits

7 6

2.5

5 4

15 3

1.5

03.17

05.16

07.15

09.14

11.13

01.13

03.12

05.11

07.10

09.09

11.08

01.08

10 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

0.5

Note: Profitability ratios are estimated via dividing one year accumulated profit by associated denominator

Note: Overdraft accounts, credit card accounts and zero-interest credits since July 2015 are excluded.

Source: CBRT (Latest Data: 03.17)

Source: CBRT (Latest Data: 04.17)

The capital structure of the banking sector remains strong. As of March 2017, the sectorsâ&#x20AC;&#x2122; capital adequacy ratio (CAR) was well above the regulatory (8 percent) and target ratios (12 percent), and the sector continued to have adequate buffers to cover possible negative shocks. Credit rating agencies' outlook related to the banking sector had a limited impact on external funding costs. On the other hand, the leverage effect driven by the recovery in credit volume and the increase in risk-weighted assets triggered by the exchange rate effects on foreign currency assets have slightly reduced the capital adequacy ratio. However, the capital adequacy ratio has reached the level of the third quarter of last year due to the stability in exchange rates, the upward trend in profitability and the changing risk weights (Chart 11). In this report period, the banking sector has been

Financial Stability Report - May 2017

able to sustain loan growth rates through capital generation as well as through foreign financing. The increase in profitability has been the main factor boosting the regulatory capital within the last one year period. On the other hand, the increase in yields on government domestic debt securities (GDDS) has limited the previous positive effect arising from the revaluation of securities (Chart 12). Chart 11

Chart 12

CAR and Core Tier 1 CAR

Changes in Items Affecting Capital

(Percent)

(12-Month Accumulated, Billions TL)

Dispersion of CAR (10 Largest Banks) CAR (Sector) Core Tier 1 Car (Sector)

Adjustment and Evaluation Items Profit, Reserves and Paid Capital Balance Sheet Capital Subordinated Debts Regulatory Capital Other Items

50 40

30 20

-10

Source: CBRT (Latest Data: 03.17)

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Source: CBRT (Latest Data: 03.17)

Global geopolitical developments and uncertainties about economic policies are perceived as the main risk factors undermining the global financial stability in the upcoming period. Other potential risk factors include the limited size and impact of the expansionary fiscal policy projected to be implemented in the US, the emphasis on possible protective foreign trade policies, policy normalization of the Fed, the possibility of a decline in the risk appetite due to the political uncertainties in the European Union and the concerns about the Chinese economy. It is assessed that the Turkish banking sector is resilient to such risks thanks to its strong capital structure, asset quality and liquidity buffers.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

I. Macroeconomic Outlook The uncertainty in global financial markets following the US presidential elections has partly abated since the last report period. During this period, the decline in global uncertainties and the recovery signals of global growth have led to strong performance of stock market indices, especially in advanced economies. On the back of the decline in inflation expectations, demand in bond markets has recently increased and long-term interest rates have assumed a downtrend again. Since the beginning of 2017, emerging economies have enjoyed portfolio inflows thanks to the increase in global risk appetite and investors seeking higher yields. The Emerging Market Currency Index has appreciated against the US dollar owing to portfolio inflows. While it remains weak, global growth has continued to recover. Moreover, despite the recent pick-up, growth rates in emerging economies, excluding China and India,

are still

below those of advanced economies.

Economic activity showed a moderate recovery in the last quarter

2016,

driven

increased

private

consumption

expenditures and strong goods exports supported by measures and incentives introduced. Meanwhile, developments in exchange rates and oil prices, and the rise in food prices caused a rise in consumer inflation. Foreign trade continued to make a positive contribution to the current account balance, despite the decline in the share of tourism in export revenues. Exports of goods and services are anticipated to increase in the upcoming period owing to the net income effect of the global economic recovery on our export partners and the market diversification performance of

our

exporters. The central government budget deficit slightly increased in the second half of 2016 and the first quarter of 2017, when fiscal policy supported economic growth. This increase is expected to be temporary and fiscal discipline will continue to be the main anchor of the Turkish economy. As a result of all these developments, risk perceptions for Turkeyâ&#x20AC;&#x2122;s assets have been on a positive trend since the second half of 2016.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

In the first quarter of 2017, economic policy uncertainty decreased. Chart I.1.1 (Index, 2012=100)

following the US presidential election has somewhat decreased

Global

USA

UK (RHA)

150

300

125

250

100

200

150

100

0 07.16

04.17

350

01.17

400

175

10.16

200

04.16

450

01.16

500

225

10.15

250

07.15

International Developments

The uncertainty experienced in the global financial markets

Economic Policy Uncertainty Indices

04.15

I.1

since the last reporting period (Chart I.1.1). The limited progress made by the new administration regarding expansionary fiscal and protectionist trade policy objectives (Box I.1.1) and a clearer picture of normalization in the monetary policy have played a role in this decrease.

The convergence of the US inflation to the 2-percent target of

Note: Indices are not comparable among themselves in terms of level.

Source: Bloomberg (Latest Data: 04.17)

the US Federal Reserve (Fed) supports the normalization process in monetary policy. The rise in inflation and the recovery in leading

Market participants foresee a softer path in the Fed's policy rate. Chart I.1.2 FOMC Members' Median Policy Interest Forecasts (Straight Lines) and Market Expectations (Intermittent Lines)

economic activity indicators have increased the likelihood that the Fed will go for two Fed rate hikes in the rest of the year. On the other hand, there exists a divergence between the long-term Fed rate expectations of the members of the Federal Open Market Committee (FOMC) and those of market participants (Chart 1.1.2).

3.5

09.16 03.17 17.05.2017

3.0

12.16 08.11.2016

More precisely, while FOMC members predict that the policy rate will

2.5

approach 3 percent in the period up to 2019, market participants

2.0

are more cautious in assessing the recovery in the economy and

1.5

expect a more moderate increase in Fed rates. In addition, the

1.0

issues related to the downsizing of the balance sheet of the Fed

0.5

remain on the agenda.

0.0

2017

2018

2019

Note: Intermittent lines indicate 30-day Fed fund futures rates.

While stock market indices show a strong performance,

Source: Bloomberg (Latest Data: 17.05.2017)

particularly in advanced economies, in the aftermath of the US After the US presidential election, particularly the advanced economy stock market indices have shown a strong performance.

demand in bond markets due to the decline in inflation expectations (Chart I.1.3 and Chart I.1.4). In almost all emerging markets, treasury

Chart I.1.3 Stock Market Indices

bond yields that increased following the US election started to

(Percentage Change, 08.11.2016-17.05.2017)

decline after January (Chart I.1.5).

Germany DAX

US S&P 500

US Dollar

presidential election, there has also recently been an increase in

UK FTSE 100 Japan Nikkei 225 Brazil IBOV Mexico IPC China Shanghai Composite

Local currency

Germany DAX US S&P 500 UK FTSE 100 Japan Nikkei 225 Brazil IBOV Mexico IPC China Shanghai Composite

-5

Source: Bloomberg (Latest Data: 17.05.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Chart I.1.4

Chart I.1.5

10-Year Treasury Bond Yields in US, Germany and Japan (Percent)

10-Year Treasury Bond Rate in Emerging Economies (Percent)

US 10 year Germany 10 year Japan 10 year

Brazil Indonesia South Africa

India Russia Turkey

2 1

05.17

11.16

05.16

5 11.15

05.17

11.16

05.16

11.15

-1

Source: Bloomberg (Latest Data: 19.05.17)

With respect to the monetary policy stance, it is observed that the decoupling in the policies of the central banks of advanced economies has continued. At the March meeting, the Fed made the

Since the beginning of 2017, portfolio flows towards emerging economies have gained momentum. Chart I.1.6 Weekly Capital Flows to Emerging Economies (Billion US Dollar, 13-Week Cumulative)

first interest rate increase in 2017 and gave the early signal of starting Bonds

the balance sheet reduction. It is foreseen that the expansionary monetary policies will continue in Japan where the inflation rate is expected to fall short of converging to the target in the mid-term, and in the EU where concerns over the financial system persist.

Equities

45 35

USA Presidential Election

Inauguration of New US Administration

25 15

5 -5

-15

search for high yield (Chart I.1.6). The demand for high-yield debt

05.17

04.17

03.17

02.17

01.17

12.16

11.16

economies have increased due to the global risk appetite and the

-25

10.16

Since the beginning of 2017, portfolio inflows to emerging

Source: EPFR (Latest Data: 10.05.17)

instruments has increased due to the leading indicators of the global economic outlook and the developments in the aftermath of the US

With increased risk appetite, credit default swap premiums in emerging economies improved.

presidential elections. Likewise, there have been significant inflows to

Chart I.1.7

global capital markets. In line with the increasing risk appetite,

CDS Premiums in Emerging Economies (Basis Points)

portfolio inflows as well as risk premiums have improved (Chart I.1.7).

Emerging Economies

450

Selected Emerging Economies

400 350

developments. On the other hand, after the elections in the Netherlands and France, the euro appreciated against the US dollar

50 05.17

02.17

11.16

0 08.16

currencies first, and then depreciated as a result of political

100

05.16

over the office (Chart I.1.8). The US dollar strengthened against major

150

02.16

emerging economies after the new administration in the US took

200

11.15

intensive portfolio inflows particularly to bond markets in the

250

08.15

against the US dollar since November 2016 was offset as of April by

300

05.15

The depreciation of the emerging market currency index

Note: Emerging economies include Brazil, Czech Republic, Indonesia, S. Africa, Colombia, Hungary, Poland, Romania, Turkey and Chile. Brazil, Indonesia and South Africa CDS premiums are used for the calculation of the average of selected emerging economies.

Source: Bloomberg (Latest Data: 17.05.17)

as the political uncertainties in the EU declined.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr The global growth, though maintaining a weak outlook,

Chart I.1.8

continues to recover (Chart I.1.9). Leading growth indicators support

Exchange Rate Indices

this judgment for advanced economies, in particular for the EU

US Dollar Index (RHA) JP Morgan EM Currency Index (8 November = 100) Brazilian Real (8 November = 100) Mexican Peso (8 November = 100) Euro (8 November = 100) 110

US Presidential Election

100

(Chart I.1.10). While it is expected that the expansionary fiscal 118

policies anticipated to be applied in the US in the upcoming term

114

will support the global growth, protectionist trade policies and

110

political uncertainties in Europe stand out as downside risks.

106 90

102

05.17

04.17

03.17

02.17

01.17

12.16

11.16

10.16

09.16

08.16

07.16

06.16

05.16

94 04.16

Growth rates in emerging economies, except China and India, have recently recovered, but still remain below the growth rates of advanced economies. There is no improvement in the

Source: Bloomberg (Latest Data: 17.05.17)

momentum of Chinaâ&#x20AC;&#x2122;s growth that was lost during the transition from the growth model based on investment and exports to the growth model based on consumption. In addition, the shadow banking and Global growth continues to recover, albeit with a weak outlook.

high indebtedness in China continue to be viewed as financial vulnerability concerns.

Chart I.1.9 Growth of Advanced and Emerging Economies (Percentage Change, Annual)

Advanced Economies Emerging Economies Emerging Economies Excluding China and India

The upward trend observed in the commodity prices since

2016 continues (Chart I.1.11). The relative recovery in the Chinese

economy and the positive signals from the US and the global growth

6 4

played a role in the rise in the general commodity index. Oil prices

have increased slightly due to the partial implementation of the oil

0 -2

production restriction imposed by OPEC member countries. In

-4 -6 123412341234123412341234123412341234

2008 2009 2010 2011 2012 2013 2014 2015 2016 Note: Advanced Economies: USA, Euro Area, Japan, UK, Canada, S.Korea, Switzerland, Sweden, Norway, Denmark, Israel. Emerging Economies: China, Brazil, India, Mexico, Russia, Turkey, Poland, Indonesia, S.Africa, Argentina, Thailand, Malaysia, Czech Republic, Colombia, Hungary, Romania, Philippines, Ukraine, Chile, Peru, Morocco.

Source: Bloomberg, CBRT (Latest Data: 12.16)

addition, the US shale oil production and the increase in general oil stocks also played a role in this limited rise in oil prices.

Several elections took place around the world over the last year and their outcomes indicate that vulnerabilities resulting from political

will

somewhat

decreasing

the

forthcoming period. However, the early election in the United

Chart I.1.10 Manufacturing Industry PMI Indices

Kingdom, elections in Germany and the EU-UK negotiations are

USA Euro Area Japan China Emerging Economies

uncertainties

considered the political uncertainties that are expected to affect financial markets. It is thought that the steps towards reducing the

Fed balance sheet, the inflation trends in advanced economies and the developments in the banking sector will play a role in the

formation of monetary policies. It is also expected that with the

04.17

10.16

04.16

10.15

04.15

10.14

finalization of the global financial reform agenda, uncertainties regarding the regulations will disappear and reforms will be in effect. The absence of any deterioration or reversal of reforms and

Source: Bloomberg, CBRT (Latest Data: 04.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr regulations, particularly in the United States, is a precondition for an open and integrated global financial system.

Chart I.1.11 Commodity Prices (Indices, US Dollar)

S&P Metals Index S&P Headline Commodity Index Brent Crude Oil (RHA)

The global economic growth and the political and financial conditions in advanced economies are among the factors that can affect capital flows towards emerging economies. In this framework,

500

400

300

200

100

macroeconomic measures to increase the resilience of the financial

05.17

11.16

05.16

structural problems gain importance.

11.15

systems of emerging economies as well as policies focusing on

Source: Bloomberg (Latest Data: 19.05.2017)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Box I.1.I

Protectionist Policies in International Trade and Possible Effects

The ever-increasing phenomenon of globalization since the beginning of the 1980s has brought about liberal policy implementations that support global and regional integration. Developments in information technology, transportation and communications have contributed to the growth of global capital flows and trade volume. With the influence of liberal trade policies, countries have become a part of "global value chains", and a structural transformation has occurred in their production activities. Throughout this time, the sensitivity of the change in international trade volume to global growth and demand conditions has increased 1. Along with the political and economic formations such as the EU, multilateral trade agreements like NAFTA (North American Free Trade Agreement) and developments such as China's membership of the World Trade Organization (WTO) have contributed to the increase in global trade flows. However, when addressed in terms of the effects on employment, income distribution and development, there are differences in the influences of pro-globalization trade policies on countries. In the aftermath of the global financial crisis, the influences of globalization on growth and employment have become more visible due to the impacts of weak economic growth and declining investments, and protectionist trade policies have been resorted to more. According to the WTO report2, the number of trade measures implemented across the G20 reached its highest level since 2009, with a monthly average of 21 between October 2015 and May 2016. In this period, 89 of 145 trade measures consisted of countervailing measures and anti-dumping practices. In recent years, protectionist policies in developed countries have started to be voiced due to the negative developments in growth and employment, as well as increased migration movements. In this period, the start of the Brexit process, referring to the departure of the UK from the EU, resulted in an increase in political and economic uncertainty, and contributed to the spread of anti-integration policies. In addition, the implementation of protectionist trade policies was among the promises of the new administration in the US presidential campaign. Also, the uncertainty in the period after the US presidential election about the implementation of these policies is considered as a downside risk for global trade flows and economic growth that are already slowing down after the global financial crisis3. Therefore, the possible effects of protectionist trade policies envisaged to be implemented by the new administration are important for the global economic outlook. It is assessed that the composition of the US foreign trade deficit is the determinant factor for the rise of protectionist trade policies. By 2016, the US had the largest foreign trade deficit with China, followed by Japan, Germany and Mexico, respectively (Chart I.1.I). On a regional basis, it is likely that the Asian economies in which the US trade deficit intensifies will be among the countries most vulnerable to possible protectionist policies.

1 Domit, 2

3 IMF

S. and T. Shakir (2010). “Interpreting the World Trade Collapse”, Bank of England Quarterly Bulletin.

“Report on G20 Trade Measures”, WTO OMC, June 2016. Global Financial Stability Report, April 2017.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

The revision of free trade agreements,

Chart I.1.I.1 US Trade Deficit Composition (Billion US Dollar)

the renegotiation of NAFTA, the introduction of customs duty with the rates of 35 and 45 percent for imports from China and Mexico, respectively, have been raised among the possible protectionist trade policies that can be put into practice by the US4. On the other hand, in the post-election period, it was decided in the first place not to put the Trans-Pacific

Source: US Bureau of Census (Latest Data: 2016)

Partnership (TPP) agreement into effect. One reason for this is the possibility that the TPP is able to contribute more to the high-skilled workforce, but is likely to increase inequality in income distribution. In addition, the removal of the paragraph related to the prevention of protectionist policies from the G20 Communique on the recommendation of the US at the G20 Finance Ministers and Central Bank Governors meeting held in Germany on March 2017 has been a significant development echoed in international markets. However, it has also been seen that there are some legal and political obstacles to the protectionist trade policies brought forward by the US administration. Delaying the draft law amending the health reform can be considered as an example of political obstacles.

Table I.1.I.1 Turkey’s Exports to the US (Billion US Dollar)

2010

2011

2012

2013

2014

2015

Total Exports

3.76

4.58

Trade Deficit

8.55

Share of Exports in Total (%) Export Growth (%, Annual)

2016

5.60

5.64

6.34

6.40

6.62

11.45

8.53

6.96

6.38

4.74

4.24

3.3

3.4

3.7

4.4

4.6

16.1

21.8

22.3

0.6

12.4

0.9

3.6

Source: TURKSTAT(Latest Data: 2016)

“Here’s Donald Trump’s Presidential Announcement Speech,” 16 June 2015, time.com/3923128/donald-trump-announcement-speech (date of access 2 May 2017);

“Trump calls NAFTA a ‘disaster,’” 25 September 2015, www.cbsnews.com/videos/trump-calls-nafta-a-disaster (date of access: 2 May2017).

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Chart I.1.I.2 Weighted Average Effective Tariff Rates Applied by the US (2015, %)

Protectionist trade policies continue to be seen as a downside risk factor for global

economic activities in reports published by

international organizations, in particular by

the IMF. In the event that these measures are

Turkey

China

World

Textiles

Ores and Metals

and developing countries, but the direct

Manufactures

will be some negative reflections on the US

Food

Chemical

put into practice, it is estimated that there

Agricultural Raw Materials

effects on the Turkish economy are assessed to be limited. Turkey's exports to the US are on an upward trend and constituted 4.6 percent of Turkey's total exports by 2016

Note: Sectoral division is based on SITC REV2.

Source: World Integrated Trade Solution (WITS).

(Table I.1.I). Although the annual growth rate of exports to the US fluctuated between 2010

and 2016, there was a decrease in the foreign trade deficit. Moreover, the average tariff rates applied by the US on imports from Turkey are already above the world averages in food and agricultural raw materials sectors (Chart I.1.2). In this respect, if the US imposes unilateral tariff increases on China in certain sectors, the competitive power of Turkey in the US market will increase relative to China. In addition, according to empirical findings5, the high income elasticity of Turkey's exports to developed countries also supports the expectation that the effects of possible US protectionist policies will be limited. However, the effects of these policies on the global economic outlook should be closely monitored in a multifaceted way.

5 Ă&#x2021;ulha

and KalafatcÄąlar, CBRT Research Notes in Economics, 2014-05.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Box

Financial Instruments International Financial Reporting Standards 9: Expected

I.1.II

Credit Loss

High loan losses of banks and other credit institutions during the 2008 global financial crisis provided a basis for a review of the “incurred credit loss” approach that caused these losses to be represented in financial statements with a time lag and to a lesser extent than they should be. After the global financial crisis, upon the invitation of G20 leaders, legal authorities of countries and regulatory bodies, the International Accounting Standards Board finalized its studies and published the “IFRS International Financial Reporting Standards – 9” document in 2014. The document comprises new regulations on the classification and risk measurement of financial instruments, expected credit loss, and hedge accounting. This box presents information on the “expected credit loss approach”, an element of IFRS 9 accounting standards that will be implemented in 2018, which enables a more prudent loan provision mechanism. Expected Credit Loss Unlike the backward-looking approach of the IAS (International Accounting Standards) 39, the IFRS 9 adopts the approach of “utilizing full range of current and previous information that may have effect on future cash flows estimation for the credit”. The expected credit loss (ECL) is a probability-weighted estimate of lifetime credit losses of a financial instrument. In other words, it is the present value of all cash shortfalls. A cash shortfall is defined as “the difference between the cash flows that are due to an entity in accordance with the contract and the cash flows that the entity expects to receive.” ECL calculations employ any reasonable and supportable information obtained from past events, current economic conditions and forecasts of future economic conditions. Although the ECL approach of IFRS 9 and the expected loss calculation of Basel III look alike, the expected loss is calculated for a time window of 12 months in Basel III. On the other hand, the “probability of default (PD)” and the “loss given default (LGD)” elements used in ECL calculations are estimated marginally for each period in the remaining maturity of a financial asset. Meanwhile, the exposure at default (EAD) is estimated according to lifetime cash flows of the asset. The present value of the expected loss is adjusted with the discount factor (DF). The ECL may be formulized briefly as: ECL = PD*LGD*EAD*DF

The ECL is reflected in financial statements through two approaches: 12-month ECL: This refers to the portion of lifetime expected credit losses that represents the expected credit losses resulting from probable default events1 regarding a financial instrument within 12 months after the reporting date (or remaining maturity if the expected lifetime of the financial asset is less than 12 months).2 The 12-month ECL is not an estimation of the cash shortfalls in 12 months but an estimation of loss calculated by weighting the expected lifetime (as long as the asset is shown in 1

If there is no reasonable and grounded information revealing default, action is taken based on the presumption that the default will take place in maximum 90 days.

TFRS 9 (2017 Version) Financial Instruments, APPENDIX A Defined Terms

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

balance sheet) cash shortfalls with the probability of default for 12 months. Lifetime ECL: This refers to expected credit losses originating from all possible default events over the expected lifetime of a financial instrument.3 The IFRS 9 has designed a three-stage model for the recognition of impairment based on changes in the credit quality.4

Table I.1.II.1 Conditions for Impairment and Stages

Stages Recognition of Expected Credit Losses

Stage 1 12-Month Expected Credit Loss

Interest Revenue

Stage 2

Stage 3

Lifetime Expected Credit Loss

Gross Carrying Amount

Net Carrying Amount

Stage 1: This stage includes financial instruments that do not show a significant increase in the credit risk or that have low credit risk on the date of initial recognition.5 For these financial instruments, the 12-month ECL method is used to determine the loan loss provision to be set aside. The interest revenue of these assets is calculated based on the gross carrying amount of the financial asset. Stage 2: The transition to stage 2 is linked to a significant increase in the credit risk. Financial instruments that have had a significant increase in the credit risk since the date of initial recognition, or in other words, instruments that are not classified as having low credit risk are included in this stage. The ECL for these assets is accounted with the lifetime expected credit loss method. Similar to the interest revenue of assets in stage 1, the interest revenue of these assets is also calculated based on the gross carrying amount. Stage 3: In addition to the conditions set out in stage 2, this stage includes financial instruments that become credit-impaired or that are in default. For these financial instruments, the lifetime ECL method is used to determine the loan loss provision to be set aside. However, the interest revenue of these assets is calculated based on the net carrying amount which is the net of loss allowance, instead of the gross carrying amount. Example: At the beginning of January 2018, a TL 100-thousand unsecured loan is extended with a maturity of 4 years at an annual interest rate of 10 percent and with a principal payment at maturity. The LGD is assumed to be 100 percent since the loan is unsecured, and the gross amount of the loan is considered to be TL 100 thousand for all periods as the principal payment will be made at maturity.

3 TFRS

9 (Version 2017) Financial Instruments, APPENDIX A Defined Terms

4 Snapshot: 5

Financial Instruments: Expected Credit Losses (2013), IFRS.

TFRS 9 (Version 2017) Financial Instruments, B5.5.22

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

1) December 31, 2018: The credit is in stage 1. The PD for the next 12-month period is estimated to be 1 percent based on the bank model. 2) December 31, 2019: Credit risk has increased significantly as it is expected that the customer will have cash flow problems due to the deterioration in economic conditions. This loan is transferred to stage 2. The PD for the remaining life of the loan is estimated to be 35 percent. 3) December 31, 2020: Since the loan has not been paid back, it is transferred to stage 3. The PD for the remaining life of the loan is estimated to be 60 percent. As this loan was in stage 2 throughout 2020, the interest revenue will be calculated based on the gross carrying amount of the loan. 4) December 31, 2021: Throughout 2021, the interest revenue will be calculated based on the net carrying amount of the loan (net of loss allowance).

Table I.1.II.2 Example: Calculation of Expected Credit Loss Reporting Date

Stages

Gross Carrying Amount

LGD

ECL

Net Carrying Amount

Interest Revenue

31.12.2018

100,000 TL

100%

100,000*1% = TL 1,000

TL 99,000

100,000*10% = TL 10,000

31.12.2019

100,000 TL

100%

35%

100,000*35% = TL 35,000

TL 65,000

100,000*10% = TL 10,000

31.12.2020

100,000 TL

100%

60%

100,000*60% = TL 60,000

TL 40,000

100,000*10% = TL 10,000

31.12.2021

100,000 TL

100%

60%

100,000*60% = TL 60,000

TL 40,000

40,000*10% = TL 4,000

IFRS 9 Turkey Implementation: Transition Process of the Banking Sector The â&#x20AC;&#x153;Regulation on the Procedures and Principles Regarding the Classification of Loans and the Provisions to be Set Asideâ&#x20AC;?, which regulates the provisions to be set aside by the Turkish banking sector in the scope of the TFRS (Turkey Financial Reporting Standards) 9, will come into force on 01.01.2018. In this regulation, the distinction between general and specific provisions is maintained. Consistent with the IFRS 9, provisions which are set aside by the approaches of the 12-month ECL and the lifetime ECL driven by a significant increase in credit risk are considered as general provisions. On the other hand, provisions which are set aside through the approach of expected credit loss due to the default of the borrower are called specific provisions. From 2018 onwards when the new regulation will become effective, banks applying the TFRS 9 will recognize general provisions at an amount as much as the 12-month ECL for the loans that do not witness a significant increase in the credit risk. When there is a significant increase in the credit risk of the loan, the provision will be set aside to the extent of the lifetime ECL and the rate of provision will also increase in proportion to the deterioration in the loan quality (Chart I.1.II.1). According to the regulation, the relevant authority may decide to give some additional time to banks for the transition to the adoption of the TFRS 9. In this case, banks which are not eligible to apply the TFRS 9, will set aside a minimum general provision of 1.5 percent and 3 percent for their standard loans and loans under close monitoring, respectively. The specific provision ratios for banks are set at the minimum levels of 20, 50 and 100 percent for the loans under third, fourth and fifth groups, respectively (I.1.II.2).

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Chart I.1.II.1

Chart I.1.II.2

Provisions for Banks Conforming to the TFRS 9

Provisions for Banks Not Conforming to the TFRS 9 (Percent) General Provisions 100

Significant Increase in Credit Risk - Beginning of Stage 2

100

Specific Provisions

Impairment

Provision Rate

40 12-Month Expected Loss

Lifetime Expected Loss

20 20 1.5

Group 1

Group 2

0 Group 3

Group 4

Group 5

Deterioration in Credit Risk Quality

Source: IASB

Source: BRSA

The expected credit loss approach, which allows for a more prudent provision mechanism in new accounting standards, is anticipated to strengthen banks' credit risk management. Determining expected credit losses according to future risks instead of realizations could mitigate the impact of possible shocks by reducing the cyclicality of the financial system. In addition, earlier recognition of credit losses for financial assets will enable banks to make sound capital adequacy assessments. In this context, it is important that the risks likely to arise in the future are evaluated in a comprehensive manner. The validation and regular review of banks' internal models will increase the effectiveness of the ECL implementation.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

I.2

Domestic Developments In the final quarter of 2016, there was an economic recovery, quarter-on-quarter.

In the final quarter of 2016, the Turkish economy grew with the support of the final domestic demand (Chart I.2.1). The rise in the final domestic demand was mainly driven by the balancing effect of

Chart I.2.1 Contribution to Growth from the Expenditure Side (Percentage Point)

Net Exports Other Final Domestic Demand GDP Growth

macroprudential measures on private consumption expenditures. In the final quarter of 2016, the negative impact of net exports to

growth remained minimal. On the back of these developments, the

negative impacts of the geopolitical risks and the developments

detrimental to net exports were significantly reduced.

12 8 4

0 -2 -4

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

03.13

12.12

-6

Despite the economic contraction in the third quarter of 2016, economic activity recovered in the final quarter of the year. Seasonally and calendar adjusted industrial production index, which increased by 0.4 percent year-on-year, partly compensated for the previous quarterâ&#x20AC;&#x2122;s loss (Chart I.2.2).

Source: TURKSTAT (Latest Data: 12.16)

Seasonally and calendar adjusted industrial production index points to an economic recovery in the last quarter of 2016. Chart I.2.2 Industrial Production Index (Seasonally and Calender Adjusted Annual Percentage Change)

were taken against the slowdown in economic activity in the third

quarter of 2016 started to be observed in the final quarter of 2016.

and job losses increased, resulting in a rise in the unemployment rate

Chart I.2.3

(Chart I.2.3). The climb in unemployment was mainly driven by the

(Seasonally Adjusted, Percent)

12.16

09.16

01.17

09.16

05.16

01.16

01.14

01.12

contribution of the employment campaign.

09.13

activity and consequently better employment data with the

05.13

indicators have been pointing to an improvement in the economic

01.13

new employment opportunities. Since May 2017, the leading

Unemployment Rate Labor Participation Rate (RHA) Employment Rate (RHA)

09.12

industry and construction sectors stemming from the weakening in

Labor Force

05.12

slowdown in economic activity and the decline in employment in

06.16

Unemployment rate increased in the second half of 2016.

09.15

In the second half of 2016, employment tendency weakened

Source: TURKSTAT (Latest Data: 12.16)

05.15

uncertainties.

03.16

been implementing to mitigate the effects of heightened global

12.15

-3

01.15

accommodative fiscal policies that emerging economies have

-2

09.15

asserted that the mentioned steps were in tandem with the

0 -1

06.15

unemployment and increase production capacity. It can be

09.14

lower

05.14

helped

03.15

implementations

12.14

loan

09.14

growth-supportive

03.14

Led by the recently announced investment stimulus package, the

06.14

The positive outcomes of the growth-supportive measures that

Source: TURKSTAT (Latest Data: 02.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

TĂźrkiyr Consumer inflation has been on the rise since the third quarter As of the final last quarter of 2016, consumer inflation has increased owing to the developments in FX and oil prices along with the rise in food prices.

of 2016 (Chart I.2.4). The rise in consumer inflation was driven by the

Chart I.2.4

cost channel, adverse weather conditions observed as of December

Price Indices (Annual Percentage Change)

effects. While rising import prices affected consumer prices via the

were the main reason for climbing fresh fruit and vegetable prices. The services sector inflation rose due to the price increases in FX-

CPI H Index I Index

rise in costs and oil prices, the escalation in food prices and low base

dependent sectors which was triggered by the climbing oil prices

stemming from currency depreciation.

9 8 7

Although the current account deficit slightly increased in the

final quarter of 2016, foreign trade contributed positively to current 03.17

11.16

07.16

03.16

11.15

07.15

03.15

11.14

07.14

03.14

11.13

07.13

03.13

11.12

07.12

03.12

account balance thanks to the significant increase in exports of goods and despite the declining share of tourism in exports (Chart

Source: CBRT (Latest Data: 03.17)

I.2.5 and I.2.6). Another favorable development for foreign trade in the first quarter of 2017 was the regional and sectoral rise of exports, specifically to the EU countries. In the first couple of months of 2017, due to domestic demand conditions and FX movements, the contribution of foreign trade to growth is expected to be positive in net terms. Another positive contribution to current account deficit would be the likely recovery in global economic activity.

Improvement in net exports has narrowed current account deficit. Chart I.2.5

Chart I.2.6

Current Account

Foreign Trade

(12 Month Cumulative Billion USD)

(12 Month Cumulative, Billion USD, Percent)

300

Current Account Deficit Current Account Deficit / GDP (RHA)

20 60

Export Import Export / Import (RHA)

250

18 16

70 200

14 40

12 10

65 150 60

8 20

100

12.16

03.16

06.15

09.14

12.13

03.13

06.12

09.11

12.10

03.10

06.09

09.08

12.07

03.07

12.16

03.16

06.15

09.14

12.13

03.13

06.12

09.11

12.10

03.10

06.09

09.08

12.07

03.07

Source: CBRT(Latest Data: 12.16)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Exports of goods are expected to rise in the upcoming

The financing of current account deficit continues to improve.

periods, depending on several factors including the outlook in the EU

40 20

performance of exports within the year. On the other hand, the rise

-20 -40

12.16

03.16

06.15

09.14

12.13

03.13

06.12

gradually rise and create an upward pressure on the current

09.11

-60 03.07

channel. Lastly, it could be argued that commodity prices can

Portfolio Investments (Net) Net Errors and Omissions

flows and geopolitical developments are likely to affect the general

of protectionist policies poses downward risks to the foreign trade

Direct Investments (Net) Other Investments (Net) Reserve Assets

100

12.10

monetary policies in advanced economies, the trend of capital

(12 Month Cumulative Billion USD)

03.10

uncertainties surrounding the global economic activity such as

Current Account Deficit Financing Items

06.09

tradersâ&#x20AC;&#x2122; market diversification performance. However, lingering

Chart I.2.7

09.08

possible oil price increases on our foreign trade partners and Turkish

12.07

economies and overall global economy, the wealth effects of

Source: CBRT (Latest Data: 12.16)

account balance in 2017.

In the last quarter of 2016, the contribution of foreign direct investments in current account financing remained unchanged compared to the previous quarter. While the share of portfolio inflows in current account financing increased, that of the reserves maintained previous periodâ&#x20AC;&#x2122;s share (Chart I.2.7).

There is a modest rise in central government budget deficit. Chart I.2.8 Central Government Budget Balance (12 Month Cumulative, Billion USD, Percent)

While the second half of 2016 was marked by a slowdown in economic growth, fiscal policy supported growth especially through public consumption expenditures. Growth responded positively to

Budget Deficit

Budget Deficit/GDP (RHA)

Kaynak: TCMB (Son Veri: 12.16) 35

2.5

2.0

1.5

25 20

1.0

was achieved again. In this framework, the central government budget deficit has risen and the central government budget

0.5

5 0

0.0 03.12 06.12 09.12 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16

the consumption and investment stimuli and positive annual growth

deficit/GDP ratio has increased after a long period of decline (Chart I.2.8). The decline in tax revenues stemming from the deceleration in

Source: Undersecretariat of Treasury (Latest Data: 12.16)

growth was partly compensated by the indirect positive impact of the stimulus packages on tax revenues. Chart I.2.9

Chart I.2.10

Exchange Rate Basket and CDS

Interest Rates (Percent)

350

4.5

300

250

3.5

200

150

2.5

100

1.5

Benchmark GDDS

03.17

10.16

05.16

12.15

07.15

02.15

09.14

04.14

11.13

06.13

01.13

Interbank Overnight Repo Rates (5 day MA) 08.12

03.17

10.16

05.16

12.15

07.15

02.15

09.14

04.14

11.13

06.13

01.13

08.12

03.12

10.11

05.11

03.12

Exchange Rate Basket (RHA)

10.11

CDS

05.11

(Basis points, TL)

Note: Exchange rate basket is the arithmetic average of the dollar and euro.

Source: CBRT and Bloomberg (Latest Data: 12.05.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

TĂźrkiyr Since the second half of 2016, there has been a positive trend in risk perceptions for the Turkish financial assets. Five-year CDS spreads kept declining due to the improved risk perception (Chart I.2.9). Despite fluctuations in the global markets after the US elections and the global uncertainties, the recent decline in CDS premiums has had a positive impact on the borrowing costs of the private sector. Moreover, as of the second half of 2016, the upward movement in the exchange rates was reversed with the support of the positive atmosphere after the Turkish referendum amending the Constitution. As a result of the tight monetary policy stance that the CBRT adopted in the face of fluctuations in FX markets, the average funding rate and GDDS yields have increased (Chart I.2.10).

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

II. Non-Financial Sector

Household financial assets are growing faster than their liabilities. This is mainly due to the fact that the savings deposit growth rate is above the individual loan growth rate. Other financial instruments that are influential in the increase in household assets are the savings in precious metal investments and the savings that are invested in the private pension system. Households' demand for private sector debt securities declined. Housing and general purpose loans began to increase as macroprudential measures for consumer loans were loosened as of September 2016. According to the regulations in force, households cannot borrow in terms of FX and at variable interest rates (except housing loans). Therefore, household liabilities do not bear any market risk arising from exchange rate and interest rate developments. The financial leverage ratio (liability / asset), which represents the relationship between households and the financial sector with respect to relative indebtedness, continues to decline due to the developments in the household balance sheet.

The real sector production volume is on an upward trend in line with the

rising exports,

the public incentives and the

improvement in investment expectations. The ratio of the sector's total financial indebtedness to GDP has been flat since 2015, which is well below the average of emerging countries. Effective measures taken by the public sector, thanks to the room for maneuver provided by the financial discipline, have made it easier for firms, SMEs in particular, to access finance. The rate of increase in the FX open position tends to decline due to the market awareness regarding the management of the foreign exchange risk and the effect of the firms' recent tendency towards TL credits instead of FX credits. The significant lengthening in the maturities of FX credits and the low level of the NPL ratio in FX loans are the factors that reduce the foreign exchange risk of the firms.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr

II.1

The household financial leverage ratio (liability/asset) continues to decline.

Household Developments

Chart II.1.1 Household Financial Assets’ and Liabilities’ Growth Rates and Financial Leverage Ratio (Annual Percentage Change, Percentage Share)

increase in the first quarter of 2017, following the slowing trend in

Real Growth Rate of Assets Growth Rate of Assets Real Growth Rate of Liabilities Growth Rate of Liabilities Financial Leverage Ratio (RHA)

30 25

Households' total asset and liability growth rates began to

2016 (Chart II.1.1). It is assessed that this is due to the increase in total

deposits on the asset side and the increase in housing and general purpose loans on the liability side. Household financial assets and

15 52 10 5

0 03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

09.13

03.14

12.13

-5

Note: The leverage ratio refers to the ratio of average financial liabilities to the average financial assets in the last 12 months. Real growth rates were calculated using the CPI.

Source: CBRT, BRSA, CMB, MKK, TOKİ (Latest Data: 03.17)

liabilities also grew in real terms in March 2017 compared to the same period of the previous year. The household financial leverage ratio (liability / asset), which started declining at the end of 2013, remained on this track in the current financial report period (Chart II.1.1). The majority of the household financial assets are composed of savings deposits and almost all of the liabilities are made up of consumer loans (consumer loans and individual credit cards). Therefore, the positive outlook of the financial leverage stemming

The growth rate of households savings deposits exceeded the growth rate of retail loans.

from the asset and liability developments is largely determined by the relative growth rates in savings deposits and individual loans. The

Chart II.1.2 Household Loans and Deposits Growth

macroprudential measures, which were loosened as of September

(Annual Percentage Change)

2016, were very effective on the deceleration in retail loan growth

Savings Deposits Retail Loans

rates since 2013 (Chart II.1.2). The financial leverage ratio has been

Savings Deposits - FX Adjusted

24 20

declining since 2014, as the savings deposit growth rate exceeded

the retail loan growth rate.

12 8

Consumer credit loan rates have remained flat whereas the

deposit interest rates have slightly increased compared to the 03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

03.13

Note: Retail loans refers to loans extended by resident banks and includes credit cards balances. (0.6 USD + 0.4 Euro) currency basket was used to adjust the exchange rate effects on savings deposits.

previous reporting period (Chart II.1.3). Accordingly, the loan-deposit spread has declined in this period.

Source: CBRT, BRSA (Latest Data: 03.17)

Table II.1.1 Household Financial Assets

Chart II.1.3

03.16

Savings Deposits and Consumer Loans Interest Rates (Percent)

Billion TL Total Assets TL Savings Deposits FX Savings Deposits - (Billion USD) Precious Metal Deposits - (Billion USD) Bonds and Bills - Public Sector - Priv ate Sector Mutual Funds Pension Mutual Funds Other Mutual Funds

Spread Loans Deposits

20 18 16 14

12 10 8 6 4

2 04.17

01.17

10.16

07.16

04.16

01.16

10.15

07.15

04.15

01.15

Note: Consumer loan rate developments include housing, general purpose and vehicle loan rate developmentss. Spread refers to the difference between loan rates and savings deposit interest rates.

Equity Securities Repo Currency in Circulation

917.3 438.1 278.9 97.2 9.4 3.3 20.2 6.2 14.1 84.1 51.6 32.5 45.9 0.4 40.2

03.17 Perc.

Share 100 47.8 30.4 1.0 2.2 0.7 1.5 9.2 5.6 3.5 5.0 0.0 4.4

Billion TL 1,056.1 487.0 322.9 88.8 17.4 4.8 17.0 6.4 10.6 102.9 64.6 38.2 50.9 0.7 57.3

Percentage Perc

Share 100 46.1 30.6 1.6 1.6 0.6 1.0 9.7 6.1 3.6 4.8 0.1 5.4

Change 15.1 11.2 15.8 -8.6 84.7 45.8 -15.9 3.7 -24.4 22.3 25.3 17.6 10.9 51.4 42.6

Cont. to Change (Point) 15.1 5.3 4.8 0.9 -0.4 0.0 -0.4 2.0 1.4 0.6 0.5 0.0 1.9

Note: Currency in circulation as of March 2017 is calculated by taking the household share in total in 2016-III Financial Accounts Report as constant.

Source: CBRT, CMB, MKK (Latest Data: 03.17)

Source: CBRT (Latest Data: 04.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr Household

financial

assets

increased

15.1

percent

compared to the same period of the previous year (Table II.1.1). The largest contribution to the asset growth came from the increase in

Chart II.1.4 Savings Deposits of Resident Households By TL and FX Breakdown (Percentage Share, Billion USD) TL Savings Deposits Share

savings deposits (10.1 points). The contribution of TL deposits to the asset growth was 5.3 points and the contribution of FX deposits was 4.8 points. Other financial instruments that are effective in increasing household assets are savings in precious metal investments and the

FX Savings Deposits Share FX Savings Deposits in USD (RHA)

105

100

55 95 50

90 45

An analysis of FX savings deposits in US dollar terms reveals that households have preferred FX savings deposits to a lesser extent

03.17

02.17

01.17

12.16

11.16

10.16

09.16

08.16

80 07.16

35 06.16

(Table II.1.1).

05.16

03.16

demand for private sector debt securities declined in this period

04.16

savings that are invested in the private pension system. Households'

Note: FX savings deposits do not include precious metals held by the banking sector. LHA in the graph shows the shares of TL and FX savings deposits in total. For example,. as of 03.17, the share of FX savings deposits is nearly 40 percent whereas the share of TL savings deposits is 60 percent.

Source: CBRT (Latest Data: 03.17)

as of March 2017 compared to the same period last year (Chart II.1.4). FX savings deposits have increased by 15.8 percent in nominal terms compared to the first quarter of 2016. The increase in savings deposits is attributed to the fluctuations in foreign exchange. FX

Chart II.1.5 Contribution of Resident Households’ Deposit Amounts to Growth by Periods (Percentage Points)

savings deposits measured in US dollars have decreased by 8.6

03.16-03.17

09.16-03.17

percent in the last one-year period (Table II.1.1). Consequently, while

Total

the share of TL savings deposits in total deposits decreased

1 Mil +

500 Bin-1 Mil

compared to the previous year due to the depreciation in TL, FX

FX Adjusted

TL 250-500 Thou

savings deposits increased slightly (Chart II.1.4).

50-250 Thou 0-50 Thou -4

While small balance FX deposits did not change, FX deposits with high balance (worth of 1 million TL and above) FX deposits

-4

Note: FX savings deposit has been adjusted for exchange rate effect with the (0,6$+0,4€) currency basket. Refers to the deposits held by residents.

Source: CBRT (Latest Data: 03.17)

increased compared to the previous report period. (Chart II.1.5). Similarly, compared to the same period of the last year, only large balance FX deposits showed increase in exchange rate–adjusted terms, while the rise in TL deposits was observed in all quantities. Swap transactions between banks and individuals are believed to be one of the major factors behind the increase in FX deposits in this period. Through these transactions, banks provide foreign currencies in exchange for TL, and they generally ask depositors to hold these

Chart II.1.6 Households’ Gold Portfolio in the Banking System and Gold Prices (1 Ton, TL (RHA))

250

Gold in Banking System Gold Price (RHA, Reverse Order)

200

600 700 100 800

offered to the holders of large balance deposits, are considered a

depositors invest and retrieve their money in TL.

Financial Stability Report - May 2017

900 1000 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

amounts generated by swap agreements, which are generally

part of FX accounts, they have the characteristics of TL deposits as

500

150

FX balances in their bank accounts. At maturity, depositors pay back FX deposits in return for principal and interest in TL. Although the

400

Note: Gold price represents the price of a Turkish Republican coin.

Source: CBRT (Latest Data: 03.17)

Central Bank of the Republic of Turkey Türkiyr The rise in precious metal investments was another notable development in the household financial assets. Gold prices and the

Chart II.1.7 Private Pension System in Turkey

amount of gold in the banking system, which were inversely related

(Billion TL (RHA))

Government Contribution to GDP (RHA) Participation's Fund to GDP (RHA) Participation's Funds per Participant (Thou TL) Number of Participants (Million)

in the previous periods, started to move in the same direction as of

September 2016 (Chart II.1.6). In this report period, the gold price

and the amount of gold held in deposit accounts rose. The increase

in gold investments is attributed to the change in households’

investment preferences as households substitute FX deposit accounts

4 3 2 1

with gold deposit accounts. 2016

03.17

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

0 2004

Note: 2016 end year GDP is used for March 2017.

The amount of savings in the private pension system (PPS)

Source: PMC (Latest Data: 03.17)

showed a dramatic increase in line with the state contribution made since 2013 to boost domestic savings. This tendency continued in the first quarter of 2017. The participants in the PPS and the participants’ funds have increased by 8.7 percent to approximately 6 million 750

Chart II.1.8 Automatic Enrollment to Private Pension System Employee's Fund (RHA, Million TL)

1.2

thousand and by 26.1 percent to 65 billion TL, respectively compared to the previous year (Chart II.1.7). The automatic

180

Number of Employers (Million)

160 140

enrollment mechanism for the PPS came into force in early 2017,

120

through which paid employees under the age of 45 started to be

1 0.8

100

enrolled in the system gradually. As of May 2017, approximately 2

0.6 80 0.4

million 900 thousand people have been enrolled in the system via

60 40

the automatic enrollment mechanism, creating a fund amount of

0.2 20

442 million TL (Chart II.1.8). The automatically enrolled participants

0 06.01 13.01 20.01 27.01 31.01 03.02 10.02 17.02 24.02 28.02 03.03 10.03 17.03 24.03 31.03 07.04 14.04

and the fund size showed a significant rise in April when the second stage covering mainly civil servants was introduced.

Source: PMC (Latest Data: 19.05.17)

Household equity securities exceeded 50 billion TL as of May 2017 (Chart II.1.9). The positive course of the Borsa İstanbul (BİST) Index was effective in this upward trend. However, it is notable that if Chart II.1.9

the household equity securities investments are deflated by the rise

BİST All Index and Household Equity Securities Portfolio (Thousand, Billion TL (RHA))

in the BIST Index, as of March 2017 the amount of investments rose by 50

Equity Securities (in real terms) BIST ALL Index Equity Securities (RHA)

85 80

75 35 70 65

60 25 55 03.17

09.16

03.16

09.15

03.15

09.14

03.14

09.13

03.13

09.12

03.12

09.11

03.11

09.10

20 03.10

Note: Equity securities (in real terms) were calculated using CPI and a constant.

Source: CBRT, Bloomberg (Latest Data: 03.17)

Thousands

3.5 percent compared to the same period of last year. In this period, the first real estate certificates were issued under the supervision of the Capital Markets Board. A real estate certificate is an instrument which divides housing projects into smaller shares and allows individuals to buy shares in these projects. In the first public offering of this instrument at end-March 2017, demands were received at a fixed price and 3.4 million certificates were issued at a price of 42.5 TL for each one. 52.7 percent of these certificates are bought by domestic individual investors.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr

Table II.1.2 Household Financial Liabilities 03.16 Billion TL Total Liabilities (Based on Type) Housing Vehicle General Purpose Indiv idual Credit Cards Asset Management Comp' Rec. Total Liabilities (Based on Counterparty) Banks Financing Companies TOKİ Asset Management Comp'

03.17

Percent age Share

Billion TL

Percent age Share

Percentage Contributions Change

to Change

441.9

100

499.9

100

13.1

161.3 16.0 167.1 85.1 12.5

36.5 3.6 37.8 19.2 2.8

190.1 17.3 186.3 89.4 16.8

38.0 3.5 37.3 17.9 3.4

17.8 8.4 11.5 5.1 34.4

6.5 0.3 4.4 1.0 1.0

441.9

100

499.9

100

13.1

406.0 10.6 12.8 12.5

91.9 2.4 2.9 2.8

453.2 13.9 16.0 16.8

90.7 2.8 3.2 3.4

11.6 31.4 24.7 34.4

10.7 0.8 0.7 1.0

Source: CBRT, TOKİ (Latest Data: 03.17)

Following the deceleration in the last report period, household financial liabilities started to increase due to the easing in September 2016 of macroprudential measures1 for consumer loans. The main determinant of the 13.1-percent growth in household financial liabilities was the rise in housing and general purpose loans which account for nearly three quarters of total financial liabilities.

General purpose loans provided by financing companies continued to increase.

Household financial liabilities to asset management companies

Chart II.1.10

increased due to the NPL portfolio sales of banks compared to the

Consumer Loans Extended by Financing Companies Based on Type (Percentage Share)

previous year (Table II.1.2).

100

03.16

06.16

Households continued to prefer banks as a major source of

09.16

12.16

funding as they did in the previous period. However, the share of

03.17

bank loans in households’ total liabilities decreased by 1.2 points (Table II.1.2). The increased diversity of consumer loans due to the emergence of new financing companies has been effective in this decline. Consumers were using financing company loans largely in vehicle purchases in the past, and since last year, they have also

30 20 10

0 Housing

Vehicle

General Purpose

Source: CBRT(Latest Data: 31.03.17)

preferred financing company loans in purchases subject to general purpose loans. General purpose loans extended by financing companies reached 3.3 billion TL as of March 2017. Accordingly, the share of general purpose loans extended by financing companies in total consumer loans was approximately 25 percent in this period (Chart II.1.10).

1For

further information, please see the table in “Annex 1: Macroprudential Regulations on Retail and Commercial Loans and Related Effective Dates”

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Average maturity of housing and general purpose loan has been increasing since the last quarter of the previous year.

Average Retail Loan Maturity

financial crisis, were loosened to some extent at end-2016. The

(3 Month MA, Month (RHA) 45

Vehicle General Purpose Housing (RHA)

110

average maturity of the newly extended general purpose and

105

housing loans lengthened due to the increase in the maturity cap for

100

general purpose loans from 36 months to 48 months and the rise in

the loan-to-value ratio for housing loans from 75 percent to 80

percent (Chart II.1.11). Another factor affecting the increase in the

average maturity of general purpose loans was the facility of

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

The macroprudential policies for consumer loans, which were intensively implemented in Turkey in the aftermath of the global

Chart II.1.11

03.13

TĂźrkiyr

restructuring of standard loans with a maturity of up to 72 months.

Note: The average retail loan maturity is calculated according to original loan maturity.

Source: CBRT (Latest Data: 03.17)

New housing loans contributed to the increase in the share of mortgaged sales in total housing sales as well as to the lengthening

Mortgaged residential sales have significant contribution to the acceleration in the housing market in recent months. Chart II.1.12 Contribution to Housing Sales Growth, Housing Loan Monthly Interest Rate and Granted Loan Ratio (Percent, Percentage Point)

contributed positively to the growth rate of house sales since September (Chart II.1.12). The improvement in financial conditions for housing loans was another factor effective in the increased housing sales. The ratio of the number of people who were granted

Mortgaged Unmortgaged Share of Mortgaged Cost Inc. Interest Rate (RHA) Granted Loan Ratio (RHA)

in average maturities (Chart II.1.12). Mortgaged house sales have

1.30

housing loans to the number of people who applied for a loan (granted loan ratio) has been above the period average since

1.20

addition to higher demand due to interest rate developments, banks

1.00

also have a higher motivation for granting loans.

September 2016. The higher granted loan ratio indicates that in

1.10

-10 0.90 -20 03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

0.80

03.14

-30

Note: The share of mortgaged represents the share of mortgaged sales in the total housing sales in the last 12 months. Granted loan ratio represents share of granted housing loans in the total application, the average of 2014-2017 (78) is indexed to 1.

Source: CBRT, TURKSTAT (Latest Data: 03.17)

Consumer credit card balances have remained flat in the recent period (ChartII.1.13). On the other hand, when the annual change in individual credit card expenditures is adjusted for the rise in general price level, there was a slight decline.

To sum up, the household financial leverage ratio continues to Chart II.1.13 Individual Credit Card Balance (Percent)

30 20

Annualized Weekly Change (13 Weeks MA) YoY Change (in real terms) YoY Change

decline in line with the household financial balance sheet developments. Household financial liabilities do not bear a significant market risk arising from exchange rates and interest rates in Turkey since regulations in force do not allow households to

borrow in FX and at floating rates (except housing loans).

0 -10 -20 -30

03.14 05.14 07.14 09.14 11.14 01.15 03.15 05.15 07.15 09.15 11.15 01.16 03.16 05.16 07.16 09.16 11.16 01.17 03.17

-40

Note: YoY Change (in real terms) were calculated using CPI.

Source: CBRT, TURKSTAT (Latest Data: 03.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

TĂźrkiyr

Box

Gold as an Investment Instrument in the Financial System and Gold Banking in

II.1.I

Turkey an Investment Tool in the Financial System and Gold Banking in the Country The traditional demand for gold in our country has urged banks to offer new savings

products. With the "Accumulation Accounts", where the savings are directed to regular gold purchases every month and the banks have paved the way for converting gold into revenuegenerating financial products by organizing gold collection days. In 2011, the Reserve Option Mechanism (ROM), a facility allowing holding a fraction of required reserves for TL liabilities in gold, accelerated this process. In this context, physical gold is transformed into gold deposits and transferred to gold storage accounts, which are then included in the Central Bank reserves within the scope of the ROM. In the financial system, there are gold products such as gold deposit, gold swap, gold credit and Gold Exchange Traded Fund (ETF). Gold ETFs, which are liquid investment instruments, are traded as stocks, have primary and secondary markets and trace gold prices. These tools are a good option for investors willing to invest in physical gold but do not want to deal with operational burdens and costs such as gold custody, clearing and insurance. Central banks hold some of the country's reserves in gold. This gold, expressed as monetary gold, is held in liquid markets, on secure accounts and in liquid gold tools. They are used as per the rules specified in the legislation of the relevant central bank or monetary authority and in line with market developments. During the last crisis in international markets, gold played an important role as a source of liquidity. While the Fed has provided countries with US dollar liquidity against gold through the swap lines it has established, some central banks (Sveriges Riksbank) have provided liquidity to their banking systems by pledging their country's gold reserves as collateral. By May 2017, the CBRT had 447.6 tons of gold reserves worth US $ 17.8 billion and accounting for 16.8 percent of total FX reserves. The amount covered by the reserve requirement accounts constituted approximately 73 percent of these gold reserves. In this context, the reserve requirement facilities that have contributed to the accumulation of gold reserves since 2011 have been diversified as follows: 1- The ROM facility allowing maintaining up to 30 percent of reserve requirements for TL liabilities in gold 2- The ROM facility allowing maintaining up to 5 percent of reserve requirements for TL liabilities in scrap gold 3- The ROM facility allowing holding up to the entire amount of reserve requirements for precious metal deposit accounts in the form of gold With the help of these facilities allowing maintaining reserve requirements in gold, the CBRTâ&#x20AC;&#x2122;s gold reserves have been strengthened, and cost and liquidity channels of the system have been supported by providing more flexibility in the banking systemâ&#x20AC;&#x2122;s liquidity management.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

TĂźrkiyr

By 21.04.2017, 25 establishments were benefiting from the facility of keeping gold for TL required reserves and 306 tons of gold, worth TL 45.7 billion in total, was established (Chart II.1.I.1).

Graph II.1.I.1

Graph II.1.I.2

Usage of ROM for Gold

Usage of Gold Facility for Precious Metals

(Percentage)

35 100

90 85

75 70

60 55

Upper bound for gold

05.17

12.16

07.16

01.16

08.15

03.15

10.14

05.14

12.13

07.13

02.13

09.12

04.12

50 11.11

05.17

12.16

07.16

01.16

08.15

03.15

10.14

05.14

12.13

07.13

02.13

09.12

04.12

11.11

Usage of ROM for gold Upper bound

Source: CBRT

Usage of gold facility for precious metals

Source: CBRT

Within the scope of the facility allowing keeping gold for precious metal deposit accounts, as of the maintenance period starting on 21.04.2017, 12 banks out of 23 banks with precious metal deposit accounts were benefiting from the facility and a total of 12 tons of gold worth TL 1.8 billion was kept (Chart II.1.I.2). In order to bring out residentsâ&#x20AC;&#x2122; gold savings into the economy, on 21.10.2016, the CBRT introduced an additional tranche of 5 percent also known as the scrap gold option, in addition to the existing facility of 30 percent allowing reserve requirements to be maintained in "standard gold" within

the

context

the

Reserve

Option

Mechanism. The aim of the new implementation

Graph II.1.I.3 Usage of ROM for Scrap Gold (Percentage)

was to draw gold savings, called gold under the

mattress, into the economy and for this purpose,

only wrought or scrap gold collected from

residents by banks as of 03.10.2016 was eligible for

this new tranche. By 07.04.2017, 2.2 tons of scrap

and 10 banks use the option. For investors, keeping gold savings under

Number of Banks Using Facilities

05.17

04.17

03.17

02.17

01.17

12.16

scrap gold option utilization rate is 6.8 percent

0 11.16

gold worth TL 322 million was maintained. The

Usage of ROM for scrap gold

Source: CBRT

the mattress has some burdens such as storage, security and opportunity cost. The opportunity cost is important for both the investor and the Turkish economy, because this type of savings has no fixed income and holding a high amount of gold

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

TĂźrkiyr

means waiving any gain from debt securities. Therefore, gold-denominated bonds and lease certificates, which are planned to be issued by the Undersecretariat of Treasury in return for scrap gold to be collected from investors, will play an important role in integrating the idle gold stockpiled under the mattress or kept with a custody service into the financial system. The mentioned issues will not only help attract gold out of the mattress but also will allow investors to earn additional income from their gold savings.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Industrial production and investment tendency are recovering.

II.2 Real Sector Developments

Chart II.2.1 Industrial Production and Investment Tendency (Seasonally Adjusted, 3-Month Moving Average)

Manufacturing Ind. Production Index Investment Tendency Capacity Utilization Rate (RHA)

140

79 78

130

120

110

leading surveys indicate that the real sector will make an upward

contribution to growth in 2017 (Chart II.2.1). The declining industrial

production and capacity utilization rates after a series of shocks in

the third quarter of 2016 have compensated for their losses and

contributed to growth above the expectations in the fourth quarter.

The investment tendency in the real sector, which had a limited

contribution to growth during 2016, has been recovering since the

100

04.13 07.13 10.13 01.14 04.14 07.14 10.14 01.15 04.15 07.15 10.15 01.16 04.16 07.16 10.16 01.17 04.17

The moderate recovery in industrial production and the

Note: The investment tendency is obtained by adding the difference between those who expressed the investment expectation for the next 12 months as up and those who expressed it as down to 100. Latest data of industrial production index: 03.17

end of the year. The Economic Coordination Committee (ECC) investment incentive packages announced in December 2016 as well as the diminishing uncertainties are believed to be influential in

Source: TURKSTAT, CBRT (Latest Data: 04.17)

this recovery. Particularly, incentives such as increasing the share of Confidence indices are increasing in all sectors.

government contribution to investment in the manufacturing industry, supporting firms' access to finance and decreasing the

Chart II.2.2 Real Sector Confidence Index

corporate

(Seasonally Adjusted, 3-Month Moving Average)

Real Sector

Service

Retail Trade

Construction

tax

rate

strengthen

the

production

volume

and

investment tendency in the sector. The prospect of recovery in

PMI (RHA)

110

100

investment expectations and the increase in foreign demand implies that the growth in the upcoming period will gain upward momentum.

04.13 07.13 10.13 01.14 04.14 07.14 10.14 01.15 04.15 07.15 10.15 01.16 04.16 07.16 10.16 01.17 04.17

The confidence of the real sector in the economic activity,

Source: TURKSTAT, CBRT (Latest Data: 04.17)

which had been on a downward trend since the second quarter of 2016, began to rise in almost every sector as of early 2017 (Chart II.2.2). The measures taken by the public sector in coordination with other regulatory authorities had a significant contribution to this

Production volume is increasing due to accelerating exports.

development. The negative economic outlook in the services and

Chart II.2.3

retail trade sectors, which were most affected by the economic

Automobile and White Goods Production Volume Annual Growth and Seasonally Adjusted Export Quantity Index

slowdown in the third quarter of 2016, showed a positive turnaround

(Annual Percent Change, 3-Month Moving Average) Automobile Production 60 50

White Goods Production

150

limited to certain sectors and has a potential to spread across all sectors. Furthermore, the campaigns for housing sales and the VAT

Export Quantity Index (RHA) 140

40 30

in the first half of 2017, suggesting that the economic recovery is not

130

refund facility in investment-oriented construction projects are supporting the construction sector.

20 120

Among the leading indicators for economic activity, the

-10

03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

-20

Source: TURKSTAT (Latest Data: 04.17)

110

100

production of automobile and white goods has steadily increased since the last quarter of 2016 (Chart II.2.3). The rapid production increase observed in the automotive sector is mostly related to the

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr production of new models and the strong export demand. The

Firm Indebtedness is increasing due to exchange rate movements.

increase in production in the last quarter of 2016 reflects the contribution of the domestic demand which was brought forward due to the Special Consumption Tax (SCT) regulation and the

Chart II.2.4 Share of Real Sector Financial Debt in GDP and Annual Growth of FX Loans (Annual Percentage Change, Percent Share)

increase in foreign sales. The rise in the output of white goods was driven by growing exports, increased housing sales and brought

External FX Share (RHA) Domestic FX Share (RHA) TL Share (RHA) Domestic FX Growth (USD) External FX Growth (USD)

forward consumption. In addition to this, the VAT and SCT reductions applied on the sales of furniture and white goods in February 2017 also prevented possible price movements due to the exchange rate

70 60

50 40

10 30

fluctuations and kept the demand in the sector alive. Within the

aforementioned incentives, it is estimated that the competitiveness

02.17

10.16

06.16

02.16

10.15

06.15

02.15

10.14

interest-free loans to contracting companies for the financing of domestic intermediate goods were introduced. With the help of the

06.14

-10

02.14

100 percent KGF collateral for exporting firms and the granting of

10.13

scope of the ECC decisions, incentives such as the loan facility with

Note: FX growth rates are annual percentage changes of FX loans in terms of USD.

Source: CBRT, BRSA (Latest Data:02.17)

and market diversity of exporting companies increased and this had a contribution to the climb in exports. The positive signals for economic growth in the EU, which is the biggest export partner of

Firms’ financial leverage is below the average of developing countries. Chart II.2.5

the sector, and the increase in the demand by trade partners due to

International Comparison of Real Sector Credit / GDP Ratio

rising oil prices will support the real sector production volume in the

(Percent, Percent Diffference)

II.2.1 Indebtedness of Real Sector Firms While the FX borrowing by real sector companies has slowed down recently, the ratio of total financial liabilities to GDP is on the rise (Chart II.2.4). The main factor in this increase is the rise in the TL value of domestic and external FX loans starting from the third quarter of 2016. Since the growth rate of the USD value of FX loans has decreased recently as opposed to the surge in their TL equivalents, the increase in the total debt has been mainly caused by the rise in the exchange rate. The recent slowdown in the use of FX loans by firms suggests that the share of firms' total liabilities in GDP will follow a steady course in the coming period. The ratio of total corporate credits to GDP was below the G20,

2016 Q3 2016 Q3 - 2011 Q3 difference (RHA) 2016 Q3 - 2015 Q3 difference (RHA)

180 160 140 120 100 80 60 40 20 0

60 50 40 30 20 10 0 -10 -20 -30

Indonesia South Africa Brazil Poland India Russia Chekia Turkey Malaysia Israel Hungary G20-Avrg. World Avrg. South Korea Developing Avrg. Chile Portugal China

forthcoming period.

Source: BIS (Latest Data: 09.16)

Financial leverage of the energy sector is at high levels due to investment projects Chart II.2.6 Share of the Loan Debts of Main Sectors in the Sectoral Value Added and Total Loan Volume (Percent, Percent Share)

2015 2016 FX Share (RHA) TL Share (RHA)

230

emerging countries and world average as of the third quarter of 2016

180

(Chart II.2.5). As of September 2016, while the developing countries'

130

average increased by about 10 percentage points compared to the

previous year, the growth rate of Turkish firms remained very limited.

After 2011, the period when global liquidity has been abundant,

-20

countries.

Energy

Manufacturing

Wholesale and Retail

Hotel and Restaurant

İnşaat

Real Estate Activities

5 Transport and Communication

Turkish firms. This rise is almost half of the increase in developing

Agriculture and Forestry

there has been a 23-point increase in the financial leverage of

Note: Debts include domestic loans and interemediated external loans via a domestic bank.

Source: BAT Risk Center, TURKSTAT (Latest Data: 12.16)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr When we look at the ratio of the main sectors' debt to the

The share of project financing loans is increasing.

value added by these sectors, we see that the energy sector

Chart II.2.7 Share of Loans Granted for Project Financing in Total Loans (Percent Share)

Energy Construction

Infrustructure Other

(electricity, gas and water resources) has the highest leverage (Chart II.2.6). The leverage ratio in the manufacturing industry and the wholesale-retail trade sectors, where FX and TL loans are used

the most, is 80 percent. The relatively low leverage level in the

4.40

4.38 4.66

4.69

1.94 1.24

1.99 1.29

6.98

7.16

12 10 8

construction sector, which makes a significant contribution to the

4.30 2.35

2.12

2.31

2.05

2.58

7.67

7.42

economic activity, is a positive indicator. Investments subject to

3.03

various incentives, such as the construction of power plants, are

6 4

considered to have an important role in the loan usage by the

8.03

energy sector that holds the highest leverage ratio. This argument is

2 0 12.14

06.15

12.15

06.16

supported by the fact that the share of loans used for investment

12.16

Note: As of the date of original allocation, it includes loans of at least 20 million USD with at least 5 years of maturity

project financing in total loans has increased in recent years (Chart II.2.7). The increase in the share of loans extended for large

Source: BAT Risk Center (Latest Data: 12.16)

investment projects that create employment as well as the fact that The use of TL loans by SMEs gained momentum. Chart II.2.8

these loans have very long maturities are positive factors with regard to potential economic growth.

Developments in SME Loans (Percent Share, Annual Percentage Change)

The share of SME loans in total domestic loans has increased since the beginning of 2017 (Chart II.2.8). With the introduction of

40 10 30

annual growth of SMEs' TL loans has gained momentum. With these

TL Share TL Growth (Right A.)

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

-10

03.14

incentives, SMEs, which are primarily influenced by economic fluctuations, have gained the opportunity to acquire working capital

FX Share FX Growth (Right A.)

Source: CBRT (Latest Data: 03.17)

The interest rate differential between the SME and large corporate loans has narrowed. Chart II.2.9

that will increase their commercial activities and to realize postponed investments.

The interest rate differential between the SME and large corporate loans has tended to decrease since the last quarter of 2016 (Chart II.2.9). The interest-free KOSGEB loans and TOBB Respite

TL Financial Costs of SMEs (Percent, Percent Difference)

Credits, which are known to be extended predominantly to SMEs,

Micro-Large Interest Differential (RHA) Large Firms Micro Firms Small Firms Medium Firms

and the guarantees provided by the KGF have played a major role

-4

downside risks in economic activity have provided substantial

03.17

discipline, various measures taken by the public sector to prevent

12.16

-2

09.16

06.16

Respite and KOSGEB loans. Thanks to the room provided by the fiscal

03.16

12.15

most advantage in terms of financing costs, in particular from

09.15

06.15

03.15

segment breakdown, it is observed that micro enterprises gained the

12.14

09.14

in this development. When SMEs are examined on the basis of

06.14

03.14

incentives such as KGF, Respite Credits and KOSGEB loans, the

support to the firms through the improvement in credit conditions.

Source: CBRT (Latest Data: 03.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr The rate of increase in the open position has slowed down.

II.2.2 FX Position of the Real Sector

Chart II.2.10 FX Open Position of Non-Financial Companies

The annual growth rate of FX open positions of non-financial

(Percent Share, Percentage Change)

FX Open P. / GDP FX Open P. / Export Annual Growth of FX Open P. (RHA)

companies tends to decrease (Chart II.2.10). The net open position of the real sector, which increased more rapidly compared to exports until the last quarter of 2016, started declining as of that

140

35 30

120 25

100

period. In addition to the strong performance in exports, the fact

that firms preferred TL loans instead of FX loans by acting cautiously

rate of open positions had also remained limited. Furthermore, the

02.17

10.16

06.16

02.16

10.15

06.15

02.15

10.14

06.14

02.14

in which similar exchange rate movements appeared, the growth

10.13

02.13

development. As a matter of fact, in the first three quarters of 2015,

06.13

due to exchange rate movements has been effective in this

Source: CBRT (Latest Data: 02.17)

market awareness raised in various platforms, primarily the Financial Stability Committee, concerning the management of foreign

The maturity of FX liabilities continues to lengthen.

exchange risk and the avoidance of FX contracts by the public

(Percent Share)

FX loans and their share continues to increase (Chart II.2.11). As for

external liabilities, loans with maturities of more than five years have

the largest share in total, with a further upward trend. The prolonged

120 percent indicate that companies do not have difficulty in

03.15 02.17

33.7

03.16

30.0

20.5

External FX 35.0

25.0 20.0

15.0

years and above constitute more than 50 percent of total domestic

15.0

4.2

15.9

Real sector domestic FX loans with original maturities of five

03.13-03.15 avrg. 03.16 02.17

40.0

14.0

Domestic FX

16.7

60 50

maturities in FX loans and the external debt roll-over ratio at nearly

55.9

position.

Maturity Breakdown of Domestic FX Loans and External FX Liabilities

11.2

important factors in putting a brake on the acceleration in the open

Chart II.2.11

12.8

sector in the transactions with the private sector are believed to be

10.0 5.0 0.0

0-1 1-2 2-3 3-5 5+ Year Year Year Year Year

Note: Domestic loans are grouped by original maturity and external debts are by remaining maturity.

Source: CBRT (Latest Data: 02.17)

accessing external financial markets. Most of the FX loans are concentrated in a small number of firms.

Of the total FX debt, 83 percent is concentrated in two thousand of the nearly 30 thousand firms having FX loans (Chart II.2.12). When examined in more details, 10 firms account for 10 percent of the total FX debt, 100 firms for 35 percent and 500 firms for 63 percent. Hence, monitoring the stock FX positions and short-term FX cash flows of the aforementioned firms, in which FX loans are concentrated, is important in terms of effective management of FX risk. In this context, it is planned to strengthen the supervision

Chart II.2.12 Distribution of FX Loans by Firm Number (Percent Share, Billion TL)

100 90 80 70 60 50 40 30 20 10 0

900 800 700

600 500 400

300 200 100 0

infrastructure for the FX risk of companies in the forthcoming period with the systemic data monitoring system which is currently in the process of being completed by the CBRT.

12+ Months Amount (RHA)

0-12 Months

Note: Firms are ranked according to their total FX loan balance (domestic plus external). There are 29,457 companies using FX loans.

Source: BAT Risk Center (Latest Data: 02.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr The NPL ratio in FX loans, one of the most important indicators

The NPL ratio in FX loans is very low.

regarding the measurement of the risk level of firms' foreign

Chart II.2.13 NPL Ratio in FX Loans

exchange debt, is at 1.5 percent level (Chart II.2.13). The FX NPL ratio

(Percent Growth, Percent Share)

generated from quarterly independent audit reports of the 10 banks

FX Credit Growth TL/USD Volatility FX NPL Ratio (RHA)

ratio. Moreover, it is observed that the effect of exchange rate

movements and loan growth developments on the FX NPL ratio is

rather limited, and the aforementioned ratio stayed in the band of 1

to 1.5 percent even in periods when the exchange rate volatility was

high.

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

Note: Compiled from quarterly independent audit reports of 10 banks with the largest FX loan receivables by making exchange rate adjustment. The 3-month volatility of the TL / USD exchange rate implied by the option prices has been used and shown in the left axis.

Source: CBRT (Latest Data: 12.16)

Firms' FX deposits have increased recently (Chart II.2.14). It is estimated that this increase has been due to the prudent attitude of the companies regarding future FX payments. When the maturity

FX deposits of domestic firms are increasing. Chart II.2.14

breakdown of FX deposits of firms is examined, the share of demand deposits within total deposits has decreased while the share of FX

Developments in FX Deposits of Domestic Firms (Billion Basket, Percentage Change)

deposits with maturities of three months and above has increased

FX Deposits Share of Deposits with 3+ Months Maturity (RHA)

with the largest FX loan volume is quite low compared to the TL NPL

14 12

rapidly. As the global and political uncertainties diminish, the expected results of the monetary policy are achieved and the volatility of the exchange rate is stabilized, it is expected that the real sector's FX demand, which is above the macroeconomic

03.17

11.16

07.16

03.16

11.15

07.15

03.15

11.14

07.14

03.14

11.13

35 30

07.13

03.13

dynamics, will return to its natural path.

II.2.2 Financial Risk Analysis of the Real Sector

Note: Tl equivalents of FX deposits were converted into basket value using weights 0.6 for TL / USD rate and 0.4 for TL / euro.

An analysis of the basic financial indicators of the publicly

Source: CBRT (Latest Data: 03.17)

Cash positions improve as financing costs increase. Chart II.2.15 Financial Indicators of Publicly Listed Real Sector Companies (Percent, Percent)

4.0

3.5 3.0

2.5

listed firms reveals that the ratio of total liabilities to assets continues to trend upward (Chart II.2.15). However, the rise in the leverage ratio has been rather limited since 2015, in line with the banks' cautious stances. The cash rate, which shows the ratio of the company's liquid assets to total short-term debt, started to increase from the second half of 2016. The rise in the liquidity position in this period when the economic activity tended to weaken is attributed

2.0 30

1.5

1.0

12.16

08.16

04.16

12.15

08.15

04.15

12.14

08.14

04.14

12.13

0.0

08.13

04.13

0.5

12.12

Financial Expenses / Assets (RHA) Total Liabilities / Assets Cash Ratio Note: Interest Coverage Ratio= EBITDA / Interest Expenses. Exchange rate-driven interest expenses have also been included in interest expesnses. Cash Ratio = (Liquid Assets + Securities) / Short Term Debt. As of the latest data, 236 real sector companies are included.

to the cautious attitudes of firms and their eagerness to protect their cash positions. Moreover, as mentioned above, it is considered that the increase in the FX deposits of companies in the period concerned also boosted their cash position. Although the ratio of companies' financial expenses to their assets rose in the last quarter of 2016, this rise remained below the increases in the second half of

Source: FÄ°NNET (Latest Data: 12.16)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr 2013 and the third quarter of 2015, where similar exchange rate movements had occurred.

Companies’ net end-of-period profits as well as their earnings before

interest,

tax,

depreciation

and

amortization

(EBITDA)

decreased in the second half of 2016 (Chart II.2.16). The decline in

Profitability of firms declined. Chart II.2.16 Firms’ Profitability (Percent Share, Percent)

net profit was due to the decrease in the EBITDA as well as the increase in financing expenses. Although the total EBITDA dropped, this did not increase the number of firms with EBITDA losses. It is foreseen that firm profitability will be even stronger in the upcoming

100

10 8

60 6

period due to the disappearing uncertainties, the significant

recovery in the economic activity, and the decrease in exchange

(Debt / Asset) above 60 percent are considered to have a low level of

solvency,

while

those

above

and

below

these

values,

12.16

08.16

04.16

12.15

08.15

04.15

12.14

Firm Share (EBIDTA>0) Firm Share (Net Profit>0) EBIDTA / Assets (RHA) Net Profit / Assets (RHA)

According to the general view in the related literature, firms with an interest coverage ratio (ICR) below 1.5 and a leverage ratio

08.14

04.14

12.13

08.13

04.13

12.12

rate volatility.

Note: EBITDA: Net Profit + Financial Expenses + Tax Expense + Depreciation and Amortization Costs. Firm share is the ratio of the number of firms with positive profits to the total number of firms in the sample. As of the latest data, 236 real sector companies are included.

Source: FİNNET (Latest Data: 12.16)

respectively, are defined as strong companies in terms of solvency. The interest coverage ratio in the majority of the sectors is above 1.5 (Chart II.2.17). Among these sectors, the food and the basic metal industry sectors can be considered as the strongest sectors with low leverage ratios. Although holdings operating in many sub-sectors

Sectors' ability to pay off debt is high.

that play a significant role in the country's economy have a

Chart II.2.17

leverage ratio of 75 percent, the fact that their ICR is close to 4

(Percent)

Sectoral Indebtedness and Interest Coverage Ratio 90 Energy

indicates that these companies have a high ability to manage their debt with operating profits. Despite the fact that the ICR of the hotel

Wholesal e/ Retail Holding

and restaurant sector, which incorporates the tourism sector, fell below 1.5, the low level of the financial leverage of the sector is believed to be a risk reducing factor. The energy sector has high leverage due to the recent large investments and the net cash flows that have not yet been generated. However, the potential risk for this sector is anticipated to be low because the loans used by the sector are long-term and cash flows will start to recover with the completion of investments. On the other hand, the high leverage in the wholesale and retail trade sector is attributed to the large working capital loans required by the sector.

-0.5 0

0.5

Metal Transport/ ProductInformatics Communic Service Chemistry ation Basic 60 Metal 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5

Textile Food Paper/ Hotel/ Restaurant Printing Construct ion Debt / Asset

ICR

30 Note: Interest Coverage Ratio= Net Operating Profit (EBIDTA) / Interest Expenses. FX rate expenses are included in the interest expenses. In the literature, the generally accepted Debt / Asset lower limit is 60 percent and the ICR value is 1.5. As of the latest data, there are 262 companies including holding companies.

Source: FİNNET (Latest Data: 12.16)

A close look at the ratio of foreign exchange open positions relative to the assets of the sectors suggests that the energy, the

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr basic metal and the wholesale-retail trade sectors had a relatively The open positions of holding companies relative to their assets are very low.

higher open position rate as of the last quarter of 2016 (Chart II.2.18). This situation in the energy and basic metal industry is explained by

Chart II.2.18 Sectoral FX Open Position / Assets

long-term FX loans required for investment projects. Also, the fact

(Percent)

that the pricing in these sectors is FX indexed and the high export

Energy Basic Metal Wholesale and Retail Hotel and Restaurant Transport and Comm. Textile Construction Service Food Chemical P. Metal Product Mineral Informatics Holding Paper/ Printing

earnings of the basic metal industry limit potential risks. With tourism revenues falling, the FX open position rate in the hotel-restaurant sector increased in 2016. An analysis of the consolidated balance 12.15

sheets of holding companies that have a significant portion of total

12.16

FX open positions reveals that the ratio of their FX open positions to assets is low. Given that the FX risks of firms operating under the subsectors of a holding company are also managed by the parent 0

Note: As of the latest data, there are 262 companies including holding companies.

considered to be lower.

Source: FÄ°NNET (Latest Data: 12.16)

When firms with FX open positions are examined, there is a

There is a strong correlation between export revenues and FX open positions.

strong correlation between export revenues and open positions

Chart II.2.19

(Chart II.2.19). Firms have been categorized in three groups; firms

FX Open Position and Export Revenues (Percent Share, Percent Share)

100 80

with no open position (1st group), firms below the median open 60

position to asset ratio (2nd group) and firms above the median (3rd

group). It is found out that the group's share in total export revenues

40 60 30 40 20 20

10 0

03.12 06.12 09.12 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16

Export Share (3. Group) Export Share (2. Group) Export Share (1. Group) Number of Firms (2. and 3. Groups, RHA) Number of Firms (1. Group, RHA) Note: Group 1 represents firms with no open position; Group 2 represents companies that are below the median value by open position / asset ratio and Group 3 is above the median. As of the latest data, 236 real sector companies are included.

Source: FINNET (Latest Data: 12.16)

company, the risk of open positions in other sectors may be

increases as its open position expands. Although the first group, which does not have an FX open position, constitutes 50 percent of the total number of firms, its share of export revenues is around 10 percent. When we look at the other two groups consisting of equal number of firms, the share of the 2nd group in total exports is around 30 percent while the share of the 3rd group with high open positions is about 50 percent. The fact that the FX open position is concentrated in firms with foreign sales and foreign exchange income stands as a factor reducing the FX related balance sheet risks of firms.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr

Box

Rediscount Credits and Amendments Made to the CBRT Regulations on

II.2.I

Rediscount In the scope of Article 45 of the Law on the Central Bank of the Republic of Turkey, rediscount

credits are extended to exporters and foreign exchange earning firms with a maturity of up to 240 days – 360 days for the exports of high-tech industrial products, the exports to new markets and foreign exchange earning services – via the acceptance of foreign exchange bills for rediscount. As the repayments of these credits, which are extended in Turkish liras through the Export Credit Bank of Turkey (Türk Eximbank) and other commercial banks, are made in foreign exchange on the date of maturity, they help boost the CBRT’s net foreign exchange reserves. The interest rate to be applied to rediscount credits with a maximum maturity of 120 days is the monthly LIBOR or EURIBOR interest rate, and the interest rate to be applied to rediscount credits with a maturity of 121-360 days is the 6-month LIBOR or EURIBOR rate. These favorable interest rates and maturities help significantly to reduce exporters’ financing costs. Hence, the rise in the number of firms that are extended rediscount credits as well as the increase in credit distribution by sectors and regions contribute to the diversification of Turkey’s export markets and export products, and improve balancing of the foreign trade by also bolstering exports and FX-earning services sectors. In view of the contribution of rediscount credits to the reduction of the current account deficit and to the strengthening of the CBRT’s foreign exchange reserves, various amendments have been made to the CBRT regulations on rediscount credits to increase the effectiveness of this facility as well as broaden its accessibility for the exporters. Within this context, the global credit limit, which was set at USD 17 billion on 23 January 2015, was raised to USD 20 billion on 26 July 2016. Out of this amount, USD 17 billion was allocated to Türk Eximbank and USD 3 billion was allocated to commercial banks. The credit limits per company have also been raised over the years. Currently, the rediscount credit limit per company is USD 400 million for foreign trade capital companies and USD 350 million for other companies. The entire limit can be used for credits with a maturity of up to 120 days whereas only up to 60 percent of the limit can be used for credits with a maturity of 121-360 days. Moreover, firms with a net sales revenue above TL 10, 15 and 20 billion in the latest financial year can utilize this credit facility up to two, three and four times, respectively, the limit set for them. There have been some amendments in the implementing regulations recently so as to increase both the credit utilization facility and the rediscount credits’ contribution to CBRT’s foreign exchange reserves. In October 2016, the requirement of a minimum of three signatures on the bills to be accepted for rediscount was reduced to two signatures. Consequently, the amount of rediscount credits extended through commercial banks has increased rapidly due to reduced credit costs while at the same time the number of firms that utilize rediscount credits has grown significantly as a result of broadened accessibility to credits by the exporters. As a matter of fact, the amount of rediscount credits extended through commercial banks, which was around USD 180 million in 2014 and 2015, reached USD 1.5 billion during the November-December 2016 period. Also, in 2016, the number of

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr firms that used rediscount credits increased by 360 to 1,670 and the number of projects for which these loans were extended rose by 14 percent to 10,415 compared to 2015. To support firms against high financial market volatility, it was decided in November 2016 that the maximum maturity of rediscount credits, which would be due by 31 December 2016, could be extended until 31 March 2017. According to the same decision, in case of payment on maturity without utilizing the maturity extension option, credit repayments could be made in Turkish liras and the Central Bank buying exchange rates prevailing on the maturity date would be applicable in these transactions. Within the context of the TL repayment facility, 37 credits with a total amount of USD 51 million were repaid as TL 171 million in December 2016. Furthermore, USD 330 million worth of 75 credits’ maturities were extended by 3 months. In February 2017, during the high FX rate volatility and consequent poor price formation period, with the intention of facilitating firms’ repayment of foreign exchange debt by safeguarding the financial stability, it was decided that the repayments of rediscount credits, which were extended before 1 January 2017 and would be due by 31 May 2017, could be made in Turkish liras provided that they were paid at maturity. The Central Bank buying exchange rates announced on 2 January 2017 would be applicable for these transactions. Within the context of this facility, 1,322 credits with a total amount of USD 3 billion were repaid in Turkish liras as of the 20 February-28 April 2017 period. Since the changes in regulations that allowed for reduced costs, extended maturities, raised limits and widened scope have made rediscount credits more attractive for companies, it is expected that the number of firms demanding credit will escalate and the credit utilization will become prevalent. Within this context, rediscount credit extensions, which amounted to USD 18.1 billion in 2016, are expected to reach USD 19 billion in 2017 (Chart II.2.I.1). Rediscount credits’ contribution to the Central Bank’s foreign exchange reserves, which was USD 15 billion in 2016, is expected to be around USD 14 billion in 2017 (Chart II.2.I.2).

Chart II.2.I.1

Chart II.2.I.2

Rediscount Credit Utilization and Outstanding Balance

Contribution to CBRT Reserves

(Million US dollars)

20,000

18,000 16,000

Utilized Amount Outstanding Balance 15,083 15,279

18,080

18,736

13,440

12,000

10,873

10,486 8,501

10,000 8,000

6,000

12,664

14,000

15,022

13,000

14,548

14,000

12,000

15,182

16,000

10,926

10,000

8,309

8,000

7,866

6,222

6,000

3,802

4,000

2,000

2012

2013

2014

2015

2016

2017

2012

2013

2014

2015

2016

2017

Note: 2017 credit utilization is 12 month cumulative data.

Source: CBRT (Latest Data: 28.04.2017)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr The rediscount facility of the CBRT has also started to be employed to promote the use of local currencies in foreign trade. To this end, it was decided in December 2016 that the CBRT would extend rediscount credits to banks via the acceptance of Chinese yuan (CNY) bills for rediscount, to finance the trade and investment activities between Turkey and China. The overall limit of these credits will be CNY 2 billion. The funding source of these credits is the fund in CNY obtained as part of the swap agreement signed between the CBRT and the Peopleâ&#x20AC;&#x2122;s Bank of China (PBOC). This credit is extended and repaid only in CNY, with a maximum maturity of 365 days and a maximum credit limit of CNY 400 million per company.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

III. Financial Sector As

result

the

fiscal

incentive

policies,

easing

macroprudential measures and interest rate developments spurring demand for retail loans, credit growth started to accelerate as of September 2016 in retail loans and as of the end of the year in corporate loans. Loan growth has been supported by developments both on the supply and the demand sides. NPL ratios have been stable thanks to the moderate increase in economic activity, the recovery

credit

growth,

and

changes

regulations

restructuring. The recovery in credit growth is expected to continue in the upcoming period with the effects of supportive measures and developments in economic activity.

Banks' resilience to liquidity risk continues. The tendency of the sector to roll-over its short-term external debt with long-term resources has largely continued in the recent period. The extension of the maturities of banksâ&#x20AC;&#x2122; external borrowing has increased the resilience of the banking sector against possible global liquidity shocks. The liquid buffers that banks can use in the face of any potential volatility in international markets are also strong enough to respond to the most negative scenario. Recently, the impact of favorable international market conditions has led to an increase in long-term bond and subordinated bond issues. Credit growth rates picked up following regulatory changes in retail loans and supportive measures in corporate loans. Chart III.1.1 Annual Loan Growth

Although the general outlook of short-term interest rate-

(FX-Adjusted, Percent)

Total Loans

Corporate

Retail

sensitive TL and FX positions are similar, the ratio of FX open positions with a maturity of up to month to total interest rate-sensitive liabilities

has been on the rise due to the recent shift towards FX deposits.

Nevertheless, the analysis show that the system's own resources are

at a sufficient level against any shocks that may arise. On the other

hand, it is observed that the banks hold reasonable FX open

positions in their balance sheets and they are rather prudent in

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

hedging these positions with off-balance sheet transactions.

Note: FX-indexed loans are included in FX loans and adjusted for exchange rate using a weighted basket of 0.3 for euro and 0.7 for US dollar. Based on stock data, annual growth rates are calculated over monthly values until March 2017, and weekly data for April.

Source: CBRT (Latest Data: 04.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr Although the profitability indicators of the banking sector displayed a flattening trend in the final quarter of the previous year, they continued their uptrend in 2015. An analysis of factors affecting

(Percent)

that

improvement in net interest income, continued austerity measures in

non-interest expenses and the positive outlook in securities, foreign

exchange and derivatives position have affected return on assets

(ROA) positively whereas the impact of the limited increase in NPLs

-2

appreciation effects in FX assets; the CAR has attained the third quarter figures thanks to stabilization of FX rates, increase in

03.17

06.16

09.15

12.14

03.14

06.13

09.12

in risk-weighted assets on the back of the recovery of TL loans and

12.11

adequacy ratios (CAR) decreased slightly due to the rapid increase

Annual Difference (RHA)

03.11

have affected ROA negatively. Meanwhile, although capital

Ratio

06.10

reveals

09.09

items

12.08

statement

03.08

income

06.07

09.06

based

Credit/GDP Ratio

12.05

profitability

Chart III.1.2

Note: The ratio takes stock of credit over the sum of monthly GDP over the past 12 months.

Source: CBRT, TURKSTAT (Latest Data: 03.17)

profitability and arrangements in risk weights Chart III.1.3

III.1 Credit Developments and Credit Risk

Annual Change in Credit Stock to GDP (Percent)

(∆ Credit)/GDP (∆ Credit)/GDP (FX-adjusted)

Credit growth rates started to recover with the revival in retail

expenditures, the partial easing in macroeconomic measures applied to retail loans and new regulations on debt restructuring. The growth in retail loans was mainly steered by the dynamics in housing

03.17

06.16

consumption

09.15

private

12.14

for

03.14

incentives

06.13

fiscal

09.12

loans,

12.11

housing

03.11

06.10

-2 09.09

III.1.1). The recovery in retail loans was driven by falling interest rates

12.08

12.05

effect registered at 16.3 percent in the last week of April 2017 (Chart

03.08

06.07

December, and total loan growth adjusted for the exchange rate

09.06

loans starting as of September 2016 and commercial loans as of

Note: The annual change in credit is reported as a ratio of flow GDP. The change in corporate FX credits takes 3-month differences of stock values to calculate the flow variable. The value is then FX adjusted using 3-month averages of CBRT buy rates. Annual values are calculated by adding up 4 quarters. FX-indexed are included in FX loans. The blue dashed line shows the long term average since 2004 of the FX-adjusted value.

Source: CBRT, TURKSTAT (Last Data: 03.17)

and demand loans. Corporate loans have started to recover across all firm sizes in TL loans with the supportive incentives and measures introduced. Following the increase in the KGF (Credit Guarantee Fund) collateral guarantee limit, loan growth was driven by public and several large-scale banks in the first months of the year, while

Credit growth rate and two-year differences in growth are high in international comparisons. Chart III.1.4 International Comparison of Credit/GDP (Percent, Percentage Points)

the recent sector-wide spread of the KGF implementation has accelerated loan growth. As a result of these developments, credit

2016 Q3 Credit/GDP 2016 Q2 - 2014 Q2 Difference (RHA) 2016 Q3 - 2014 Q3 Difference (RHA)

175 150

20 15

growth increased faster than GDP growth and credit/GDP ratio

125

surpassed the 70% level (Chart III.1.2). Annual net credit usage from

100

Mexico

Indonesia

Czechia

USA

Russia

Poland

India

Turkey

Brazil

-5 S. Africa

0 Japan

III.1.3).

25 Thailand

neutral series has tended towards its long-term average (Chart

China

the banking system has also accelerated, while the exchange rate-

Note: Data covers all private non-financial sector credit, with the latest data available from 2016Q3. The dashed line marks the zero line for the RHA, the two year differences are calculated between the second and third quarters of the years indicated.

Source: BIS (Latest Data: 09.16)

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Türkiyr In the third quarter of 2016, bank lending to non-financials as a share of GDP in Turkey and the annual change in this ratio have been higher than those of peer developing countries. It is expected that Turkey will rise higher in this ranking on the back of the recovery While TL corporate loans are increasing across all firm sizes …

both in consumer loans and corporate loans (Chart III.1.4).

Chart III.1.5

Both

Annual Growth in TL Corporate Loans by Firm Size (Percent)

Small Large

supply

and

demand-side

dynamics

have

been

influential in the recent recovery in bank lending. The increase in

Medium Total

demand was mainly driven by the decline in housing loan interest rates,

changes

macroprudential

measures

supportive

borrowing, sectoral measures and fiscal policy incentives. Banks 20

have kept credit supply standards stable compared to the previous period in the face of Treasury support and individual credit

incentives. It is anticipated that the positive impact of the 03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

Note: FX-indexed loans are excluded. Micro and Small SMEs are grouped together under the Small heading.

mentioned factors on credit demand and supply developments will continue in the upcoming period.

Source: CBRT (Latest Data: 03.17)

…Domestic-sourced FX corporate loans are declining in tandem with rising foreign exchange rates.

III.1.1 Corporate Loans

Chart III.1.6 Annual Growth in FX Corporate Loans by Size (FX-adjusted,Percent, TL)

Total Foreign-Sourced Small Medium Large Total Domestic-Sourced FX Basket

Since December 2016, corporate loans have accelerated

their growth rate with incentives such as KOSGEB’s interest-free loan 4

support, TOBB’s low-interest Respite Loan and Treasury-backed KGF

guarantee.1 The growth rate of total corporate loans adjusted for

exchange rates, which was recorded at 14.6 percent in March,

-5

matching previous year’s rates (Chart III.1.1). The growth rate of TL

-15

credits was 22.4 percent with a significant pick up across all scales,

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

Note: Total foreign sourced FX credit growth takes the foreign FX loans and other FX liabilities of all non-financials, excluding foreign branches and affiliates of domestic banks, in USD. FX-indexed loans are included in the total and figures denominated by size. Micro and Small SMEs are grouped together under the Small heading. The weighted FX basket uses weights of 0.3 for euro and 0.7 for the US dollar.

Source: CBRT (Latest Data: 03.17)

including SMEs, with the introduction of the KGF collateral guarantee stimulus (Chart III.1.5). Simultaneously with the increase in foreign exchange rates, small and medium-sized firms' domestic FX loans continued to shrink and large-scale FX loans continued to increase albeit at declining rates. The decline in large-scale firms’ use of FX loans, which make up 84 percent of all FX loans, was the chief determinant in the general course of total FX credit growth, while

Details of which can be found in Special Topic IV.2 titled Measures on Corporate Sector’s Access to Finance. 1

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr domestic FX loans growth decelerated and registered no growth in March (Chart III.1.6). Chart III.1.7 Corporate Loan Interest Rates and Spreads

In the past, when growth rates of domestically sourced corporate FX loans decelerated, foreign-sourced FX loan growth

(4-weekMA, Percent)

Spread (RHA) TL Loans TL Deposits 5

Spread FX Loans FX Deposits

rates would increase, as firms met their FX loan needs by shifting sources from domestic to foreign. Nevertheless, currently, we see a

slowdown in the growth rates of all FX loans with both domestic and foreign origin. As this development is synchronous with the increase

04.17

08.16

12.15

04.15

08.14

04.17

08.16

12.15

04.15

12.13

domestic usage with foreign resources (Chart III.1.6).

08.14

towards TL loans rather than FX loans instead of replacing their

12.13

in exchange rates, it is evaluated that the corporates have tended

Note: Overdraft accounts and credit cards, as well as loans with zero interest starting from July 2015 are excluded.

Source: CBRT (Latest Data: 28.04.17)

Interest rates on loans are an important factor affecting demand. The rise in TL loan growth despite the increase in loan rates and the rise in the difference between loan rates and deposit rates

Chart III.1.8

shows that the recent incentives have had a significant impact on

(Net Percent Change)

Contributions to Corporate Loan Supply

loan demand. In this period, banks increased their TL and FX deposit rates due to their funding needs for loans and this pushed TL loan rates up, while the gap between FX loan rates and FX deposit rates narrowed (Chart III.1.7). The rapid increase in demand for TL loans

Expectations Regarding General Economic Activity Industry or Firm-Specific Outlook Risk on the Collateral Demanded

-15 -30 -45

amid a rise in depositors' FX preferences urged banks to intensify currency swap transactions with the aim to obtain TL resources. The fact that companies that have used FX loans in the past have

-60 -75 -90 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

reduced their FX loan demand in the current period when the FX loan-deposit gap narrowed should be evaluated based on the

Note: The quarterly Survey asks respondents to compare the current quarter to the previous. Zero is the neutral state indicating no change.

exchange rate developments rather than cost factors.

Source: CBRT (Latest Data: 03.17)

According to the Bank Loans Tendency Survey, in the first quarter of 2017, on the back of fiscal policy incentives, banks relaxed the standards that they applied to the SME loans compared

Retail loan growth has accelerated especially due to housing and general purpose loans. Chart III.1.9 Annual Growth in Retail Loans (Percent)

standards remained flat. The credit supply standards have been

loosened in TL loans, both short and long term, but tightened in FX

loans. The developments in credit supply conditions coupled with

the decreasing credit demand played a role in the deceleration of

-5

FX loan growth. While banks expect credit standards to remain

-15

largely the same over the next three months, FX credits have also

-25

differed in this benchmark, as nearly half of the surveyed banks reported that the tightening in FX loan standards would continue. As

Financial Stability Report â&#x20AC;&#x201C; May 2017

General-purpose Housing Vehicle Credit Card

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

to the last quarter of 2016. In loans given to large enterprises, the

Source: CBRT (Latest Data: 03.17)

Central Bank of the Republic of Turkey Türkiyr expectations for general economic activity and industry-specific risk perceptions played a tightening role in standards as they did in the

Chart III.1.10 Retail Loan Lending Rates

last two years, the risk perception of collaterals has differed from this

(4-week MA, Percent)

trend owing to the stimulus packages (Chart III.1.8).

General-Purpose Housing Vehicle

19 18 17

III.1.2 Retail Loans

16 15 14 13

In March 2017, retail loans grew by 11.9 percent, due to the

decline in housing loan rates observed since August 2016, and the

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

Source: CBRT (Latest Data: 28.04.17)

increase in retail loans since September driven especially by housing and general purpose loans owing to the changes made in macroprudential regulations which increased credit and installment facilities for borrowers (Chart III.1.1, Chart III.1.9 and Chart III.1.10). The annual growth rate of vehicle lending by banks, which remained

Chart III.1.11 General Purpose Loan Weekly Growth Rates

stable despite the rising market share of financing companies can

(4-week MA, Annualized Percent)

2008-2015 Avg. 2016 Maturity Cap Change 2016 2017

be attributed to consumption brought forward by the impending special consumption tax (SCT) increment at the end of 2016, and to

the base effect in 2017. The increase in credit card growth rate, which stemmed from the base effect in the past several months, was

replaced by a horizontal growth rate.

-15 Ç1

Ç2

Ç3

Amid falling interest rates on housing loans, demand

Ç4

Note: The maturity cap change is shown to include the week it took effect in.

remained robust and housing loans continued to be the fastest

Source: CBRT (Latest Data: 03.17)

growing type of retail loan. Within the scope of the amendments to

General-purpose loan maturities are shifting in favor of the 37-48 month bracket.

limit applied to housing loans was raised to 80 percent from 75 percent in September 2016 which contributed to the housing loan

Chart III.1.12 General-Purpose Loan Maturities

growth following the fall in interest rates.1

(Stock, Percent) 1-12 Months 25-36 Months 48+ Months

existing macroprudential measures, the maximum loan/value ratio

13-24 Months 37-48 Months Maturity Cap Chg

The growth rate of general-purpose loans, which lagged

60 50

behind its long-term average, caught up with the average growth

rate in the last quarter of 2016 with the expansion of the maximum

maturity limit from 36 months to 48 months and the simultaneous

10 04.17

12.16

08.16

04.16

12.15

08.15

04.15

12.14

08.14

04.14

12.13

08.13

04.13

12.12

Note: The maturity cap change in 2013 limited the maturities to 36 months. The change at the end of 2015 removed the cap for education loans, and in 09.2016, the 36-month maturity cap was increased to 48 for all general-purpose loans. The sharp movements in the beginning of 2015 and 2016 are due to changes in definition and coverage. As general-purpose loans and “other” types of retail loans not classified elsewhere are reported together since 2015, they are graphed together for the entire duration of the Chart. The maturity cap changes are shown to include the weeks they took effect in.

According to the amendments made to the regulations regarding the credit transactions of, and credit cards issued by banks on 27 September 2016; The maturity cap for general-purpose loans, while retaining some exceptions, has been raised to 48 months and current balances on performing loans are allowed to be restructured with maturities up to 72 months. If this restructuring requires a new credit to be issued, the maturity is again limited by 48 months. The loan-to-value ratio for housing loans or loans with housing as collateral other than vehicle loans has been increased from 75 percent to 80 percent. With the exclusion of various consumption items, the number of installments in retail and corporate credit card spending and cash withdrawals has been increased from 9 to 12 months, and as in general-purpose loans, current balances on performing loans are allowed to be restructured with maturities up to 72 months. 1

Source: CBRT (Latest Data: 28.04.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr decline in interest rates (Chart III.1.11). At the beginning of 2017, the gradual rise in interest rates brought the growth rate below the

Chart III.1.13 Credit Standards and Economic Outlook (Percent, Net Percent Change)

average, but the growth path has continued to be stronger compared to the previous year. In this period, the incentives for

Demeaned Unemployment Rate Bank Loans Tendency Survey Standards (RHA) 30 1.5 20

consumption of domestic appliances and furniture, and the additional demand created in these consumption items due to the

0 -10

increasing demand for housing have created a favorable effect on

breakdown, the share of loans with maturities between 25-36 months

-20 -1.5

-30 -40 -3

which had been steadily growing, lost ground to loans with maturities between 37-48 months and this confirms that the regulatory change played a major role in credit developments (Chart III.1.12).

Household

-50 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2

the demand for general-purpose loans. In terms of loan maturity

2013

2014

2015

2016

2017

Note: The standards shown are only for general-purpose loans. The zero line shown in dashed marks is the neutral level; values below indicate tightening standars, values above indicate easing standards. Seasonally adjusted monthly laborforce statistics are used to calculate the unemployment rate over 3-month periods which is then demeaned. As the latest data available is for February, Q1 data for 2017 shows values calculated for the first two months only, indicated with a marker.

Source: CBRT, BRSA (Latest Data: 03.17)

indebtedness

and

indicators

general

economic activity will be influential on the credit risk outlook of retail

Chart III.1.14 NPL Ratios (Percent)

5.0

ratios have been declining since 2014. Consumers’ debt service

4.5

opportunities have become more favorable as economic activity

4.0

revived since the beginning of 2017, the credit standards that had

3.5

been tightening for over a year were held stable, and longer term

3.0

and installment opportunities were introduced. These developments

2.5

are expected to have a positive effect on individuals' loan

2.0

Total

Corporate

Retail

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

loans (Chart III.1.13). As stated in Section II.1, household leverage

repayments.

Source: BRSA (Latest Data: 03.17)

Chart III.1.15 Components of NPL and Their Contributions to the Monthly Growth Rate of NPL (Billions TL, Percent and Percentage Points)

Additions Receiving Write-offs

ratio has stabilized slightly above 3 percent (Chart III.1.14). Growth in

retail loans especially driven by housing and general-purpose loans

ratios at relatively low levels favors the credit risk outlook of corporate loans (Chart III.1.16).

5 0 -5

-10 Additions Receiving

Write-off NPL Growth

Source: CBRT (Latest Data: 03.17)

Financial Stability Report – May 2017

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

(Chart III.1.15). Meanwhile, the flat movement of large firms' NPL

03.15

12.14

companies contributed to the decline in NPL ratios in retail loans

12.13

and the exit from assets due to portfolio sales to asset management

Balance Under Close Monitoring

09.14

packages and the moderate recovery in economic activity, the NPL

03.14

Thanks to the positive contribution of the credit stimulus

06.14

III.1.3 Non-Performing Loans

Central Bank of the Republic of Turkey Türkiyr Turkey's NPL ratio in 2016, and the change in the ratio over the last two years are close to those of peer developing countries (Chart

Chart III.1.16 Corporate NPL Ratios

III.1.17).

(Percent)

Total

Large

SME

5.0 4.5

Corporate loans NPL ratio, registered at 2.9 percent in March

4.0

2017 while NPL ratios differed across firm sizes (Chart III.1.16). As

3.5

large-scale corporations have the largest share of loans in volume,

3.0

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

of SME loans, which has been increasing since 2014, registered at

12.14

1.5 09.14

on the aggregate NPL ratio of corporate loans. Meanwhile, NPL ratio 06.14

2.0

03.14

their NPL ratio, which was stable at about 2 percent, was influential

12.13

2.5

around 5 percent. Recent loan incentives targeting especially the SME segment through KOSGEB, the Respite Credit and KGF

Source: CBRT (Latest Data: 03.17)

guarantee schemes are expected to be effective in improving the

The NPL ratio is relatively moderate in an international comparison.

NPL rates with the reviving economic activity along with a limited debt rollover effect. As stated in Section II.2, corporate leverage

Chart III.1.17 International Comparison of NPL Ratios and Differences (Percent) 10

ratios have recently increased, however maturities have continued 6

2016 Q4 NPL Ratio 2016 Q4-2014 Q4 Difference (RHA) 2016 Q3-2014 Q3 Difference (RHA)

to lengthen. As noted in the April Inflation Report, the expected steady course of economic activity as well as developments in consumption demand and credit access will be decisive in the

6 2 4 0

USA

Japan

China

Mexico

Thailand

S. Africa

Indonesia

Turkey

Brazil

Poland

Czechia

India

-2 Russia

course of the NPL rates of both SMEs and large-scale enterprises in the coming periods. It is expected that the developments will positively affect the firm revenues and therefore the firms' debt turnover capacities.

Note: The dashed line marks the zero line for the RHA, the two year differences are calculated between the third and last quarters of the years indicated. As Japanese data is not available for Q4, the bars show values for 2016 Q3 and the two-year differences are taken for 2016 Q3 and 2016 Q1. Data not yet reported for Brazil, Czechia, China, Mexico and Thailand have been obtained from national sources, with monthly data averaged for Brazil.

Source: IMF-IFS, BRSA, Banco Central do Brazil, Banco de Mexico, Bank of Thailand, Czech National Bank, China Banking Regulatory Commission (Latest Data: 12.16)

While corporate NPL ratios diverge in terms of currencies, developments in the aggregate ratio are largely driven by TL loans (Chapter II.2). The flat course of FX loan NPL ratios confirms that firms are resilient to exchange rate shocks.

Table III.1.1

Corporate NPL ratios differ across sectors, as they do on a

Sectoral Breakdown of NPL Ratios (Percent) 03.16 NPL

03.17 NPL

Manufacturing Industry

3.1

Wholesale and Retail Trade

3.6

Construction

Percent Chang

Share of Credit

3.6

17.8

24.7

4.1

15.4

20.5

4.1

3.6

-12.7

11.5

Energy (Electricity, Gas, Water Res.)

1.1

0.5

-55.2

9.4

Transport., Inventory, Communicat.

1.6

1.8

9.4

7.7

Real Estate, Renting,Management

1.1

-1.5

7.3

Agriculture, Livestock, Forestry

2.4

2.8

19.0

5.9

Hotels and Restaurants

2.3

-1.8

4.4

Mining and Quarrying

4.4

2.4

-44.7

1.8

Note: Sectoral breakdown is based on the loan purpose indicated at the time of application. The shares are calculated excluding retail loans and the financial sector, and the selected sectors represent 93% of real sector loans.

Source: BRSA (Latest Data: 03.17)

basis of scale and currency. NPLs in the manufacturing industry and the wholesale and retail trade sectors, which together constitute approximately half of the corporate sector’s credit utilization, play an important role in the increase in the total corporate NPL ratio (Table III.1.1). It is estimated that these sectors will be positively affected by forthcoming developments in economic activity fueled by both domestic and international demand. There has been a noteworthy decline in the construction sector NPL ratios following the recovery in housing loans.

As for tourism sector NPLs, the

improvement observed in the sector’s NPL ratios, despite the difficult

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr Chart III.1.18

period that the tourism sector is going through, can be attributed to restructuring in loans and inclusion of the sector in the recent

NPL Ratios in Retail Loans (Percent)

General-Purpose Vehicles Credit Cards Housing (RHA)

incentive programs. It is estimated that restructuring opportunities also played a role in the improvement in the energy sector NPL ratios.

2.0

1.5

1.0

0.5

0.0

third quarter of 2016 (Chart III.1.14 and Chart III.1.18). The NPL ratio in housing loans, which is a stable balance sheet item in terms of both its collateral structure and its loan-to-value measures introduced as a part of the macroeconomic policy framework, remained flat at 0.5 percent in the last two years.

NPL ratios in vehicle loans and personal credit cards fell by a limited amount, to 3.3 and 8.0 percent, respectively, in March. Growth in credit card purchases repaid in installments showed a

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

retail loans have shown a downward trend across all types as of the

12.13

Recently, with the recent growth in retail loans, NPL ratios in

Source: CBRT (Latest Data: 03.17)

Balances with installments are gaining ground. Chart III.1.19 Growth in Personal Credit Card Balances and Installment Share (Percent)

With Installments Total CC Reg CC Inst. I CC Inst. III

Without Installments Installment Share CC 9 Months CC Inst. II CC 12 Months

60 35 10

limited increase, offsetting the horizontal trend in the last year on the -15

back of the increase in the maximum number of installments and the -40

As a result of this limited variation between installment and noninstallment balance growth rates, the increase in the share of installment balance to the total credit card balances did not last long and the rate has shifted back to levels observed before the regulatory changes.

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

facility allowing long-term structuring of existing loans (Chart III.1.19).

Note: The changes in the relevant regulations are respectively: in 2013, among other changes minimum payments were linked to card limits and new card limits to income. In February 2014, the number of monthly installments were limited to 9, and jewelry, telecommunications, food, and petroleum expenses were exempted from the right to installments. The 1st regulation removed the right to installments for gift cards and cheques; the 2nd brought about 4 months of installments to jewelry; the 3rd extended household goods, furniture and educational expenses to 12 months of installments. In September of 2016, the maximum installment numbers were extended to 12, and in addition to the existing exceptions, electronics and computer spending was limited by 6; airline, transportation, travel agency, hotels, health and social services, health products, club and association membership and tax payments were limited to 9months, and direct sales, sales abroad, and cosmetic and office supplies spending were exempted from installments.

Source: CBRT (Latest Data: 03.17)

NPL ratios of general-purpose loans decreased to 5.8 percent in March owing to the regulatory amendments made in September

Chart III.1.20 New General Purpose Loans and the Survey (Average Risk Group)

restructuring of existing balances with long maturities are expected

5.5

Representative Risk Group Default Weighted Group (RHA) Bank Loans Tendency Survey Standards (RHA) Bank Loans Tendency Survey Demand (RHA) 14

to ease credit customers’ debt service by reducing their monthly

5.3

obligations. This development will increase current and future

5.1

payment rates, and reduce the rate at which said loans become

4.9

NPL. Another factor that will affect the influence of these

4.7

developments on NPL ratios are the developments in credit

4.5

2016. The amendments extending the maturity cap and allowing

11 10

9 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2

standards that banks employ in assessing credit applications. According to the Bank Loans Tendency Survey, standards remained stable in the first quarter of 2017, ending the tightening of almost a year (Chart III.1.20). Standards are expected to stay flat and demand is expected to increase in the second quarter of 2017.

Financial Stability Report – May 2017

2013

2014

2015

2016

2017

Note: Standards and demand values from the Survey are only for general-pupose loans. These values, which were also graphed in Chart III.1.14 as net percent change, are rescaled in this Chart to fit the risk group range on the RHA. The dashed zero line shows the neutral point for the Survey. Values above are easing and below are tightening. 2017 Q2 values for the Survey are expected values and are shown in dashed lines. The representative and default weighted risk groups show a plain average and default probability weighted average of RLS groups for general purpose loan customers.

Source: Credit Bueau of Turkey (KKB), CBRT (Latest Data: 03.17)

Central Bank of the Republic of Turkey Türkiyr With the increase in the maturity cap in general-purpose General-purpose loan maturities are longer across all RLS brackets. Chart III.1.21 (Percent)

prior to the 2013 maturity cap announcement, they have recently

Maturity Cap Change 1 - 740 741 - 860 860 - 1100 1101- 1460 1461+

(Chart III.1.21). Although in the period immediately after the amendment new loan maturities registered at levels close to those

General-Purpose Loan Maturities by RLS 50

loans, maturities have increased in all Retail Loan Score (RLS) groups

begun converging to pre-September 2016 levels across all RLS groups. This development shows that the increase in maturities

following the amendment was not a reflection of a permanent

change in individuals’ consumption habits or banks' maturity

practices, but merely a short-term phenomenon.

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

Note: Group means are calculated using the 11 RLS groups for scores between 1 and 1900.

Source: KKB (Latest Data: 03.17)

With the amendments to regulations, maturities have lengthened and households’ debt service capacity has increased. Therefore, both loans that have been already extended, through new restructuring facilities that have been introduced, and loans to be extended in the forthcoming periods are expected to have lower

Chart III.1.22 General-Purpose Loans Vintage Curves

NPL conversion rates. As a matter of fact, according to the vintage

(Percent)

analysis, general-purpose loans issued as of the first quarter of 2015

keep performing better in terms of asset quality every quarter (Chart 3

III.1.22). It is expected that the improved macroeconomic outlook, increased employment opportunities in the current period, and

2011

2012

2013

2014

2015Q1

2015Q2

2015Q3

2015Q4

2016Q1

2016Q2

2016Q3

2016Q4

longer loan maturities will lead to a favorable NPL outlook for retail loans in the coming period.

Q16

Q15

Q14

Q13

Q12

Q11

Q10

Note: The vintage analysis reports NPL ratios cumulatively in the quarter following the issuance of a loan.

Source: CBRT (Latest Data: 03.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

III.2 Liquidity Risk The resilience of banks to liquidity risk continues. The maturities of non-core funding items have lengthened due to the measures in force and have increased the resilience of the banking sector against possible global liquidity shocks. The tendency of banks to roll over their short-term external debt with long-term resources has continued to a great extent in the recent period, implying that there is no significant change in the sectorâ&#x20AC;&#x2122;s access to external resources. In addition, the Reserve Option Mechanism and FX required reserves, which provide room for maneuver for banks to cover FX liquidity shocks even under the most negative scenarios within the one-year window, and the Liquidity Coverage Ratio (LCR), which enables banks to keep their 30-day windowed liquidity positions in the safe zone, strengthen the sectorâ&#x20AC;&#x2122;s long and short-term liquidity positions. Foreign exchange denominated issuances, which are relatively more sensitive to global liquidity developments, have been stimulated by the expectation that the monetary policies in advanced country central banks will continue to support liquidity conditions and the increase in the global risk appetite. In light of these developments, it is envisaged that liquidity constraints will not play a binding role for the banks in the near future.

Banks are sturdy against short-term liquidity shocks. The LCRs,

The sector's LCR calculated for the total is well above the legal limits.

which show how banks can meet 30-day net cash outflows of high-

Chart III.2.1

quality liquid asset stocks, are well above the legal lower limits.

(Percent, 4-Week Moving Average)

Quantiles of Banks by Total Liquidity Coverage Ratio

(Chart III.2.1 and Chart III.2.2).1 The LCRs of the sector, which are calculated for both the total and FX, currently meet the legal lower

250

25th Percentile 50th Percentile Sector

75th Percentile Legal Ratio 250 230

limit of 100 percent for the total amount to be reached in 2019 and

210

190

80 percent for FX. The development of the ratios of the banks in the

170

150

25th, 50th and 75th percentiles from the smallest to the largest,

130

110

respectively, indicates that all banks satisfy the legal limits by a significant margin. The foreign exchange and gold reserves and the required reserves of banks held within the ROM as well as their

01.14 04.14 06.14 08.14 10.14 01.15 03.15 05.15 08.15 10.15 12.15 03.16 05.16 07.16 09.16 12.16 02.17 04.17

230

Note: (1) Excluding development and investment banks. Based on non-consolidated reportings. These quantiles represent the banks in the 25th, 50th and 75th percentiles, respectively, from the smallest to the largest.

Source: CBRT (Latest Data: 05.05.16) The BRSA applies legal lower limits for LCRs, which aim to keep the short-term liquidity position of the banks in a safe zone. Since 2014, the BRSA has requested from the banks to calculate their LCR and has set the legal rate as 60 percent for the total and 40 percent for the foreign currency (FX) as of January 1, 2015. It is stated that these legal limits will be increased by 10 points each year and that in 2019, the level of 100 percent and 80 percent will be implemented as legal ratios for the total and FX, respectively. 1

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr securities portfolios constitute a significant portion of the high-quality liquid asset stock and limit the liquidity risk of the sector.

Chart III.2.2

The share of non-core liabilities in total resources has been

Quantiles of Banks by FX Liquidity Coverage Ratio (Percent, 4-Week Moving Average)

25th Percentile 50th Percentile Sector

300

following a fluctuating but flat course since the end of 2014. The 75th Percentile Legal Ratio 300

270

240

210

180

150

120

01.14 04.14 06.14 08.14 10.14 01.15 03.15 05.15 08.15 10.15 12.15 03.16 05.16 07.16 09.16 12.16 02.17 04.17

270

Source: CBRT (Latest Data: 05.05.16)

resources provided through bonds issued in foreign markets and the debts obtained from banks constitute about 60 percent of non-core liabilities. The TL equivalent amount of the funds provided from foreign sources, which constitute a significant part of non-core resources, has increased by a limited amount due to exchange rate movements during the last report period. In this framework, the share of

non-core

liabilities in total resources

has

also increased

moderately. Domestic non-core sources consist mainly of repo transactions, bank borrowings and bond issues. In this period, domestic issuances followed a flat course. With the effect of the CBRT's liquidity policies, there has been a significant contraction in the domestic repo funding, while the debt to banks has increased by the same amount.1 The share of domestic funds in total foreign resources did not show a significant change as a result of the overall flat course of borrowing and repo funding (Chart III.2.3).

The share of non-core liabilities in total resources remains flat. Chart III.2.3

which the loans having the largest share in banksâ&#x20AC;&#x2122; illiquid assets are funded with deposits, maintains its flat course. Being one of the key

Ratio of Non-Deposit Funding to Funding Sources (Percent)

The Loan/Deposit ratio (L/D), which represents the extent to

indicators of the long-term liquidity position of the banking sector, Domestic Foreign Total (RHA)

the L/D ratio approximated 120 percent by the end of 2014 and has

since assumed a flat course (Chart III.2.4). Deposit growth being

close to credit growth is a factor supporting financial stability. The

change in the L/D ratio in terms of currencies may vary depending

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

on various reasons, in particular the exchange rate movements. In

Source: CBRT (Latest Data: 03.17)

fact, during the last report period, depositors' preferences for FX deposits

strengthened.

Triggered

the

increased

market

awareness of the management of foreign exchange risk on various platforms, the foreign exchange credit demand weakened and consequently the FX L/D ratio declined. On the other hand, the TL L/D ratio moved upwards with the revival in TL loans, which had also With the announcement made on January 10, 2017, banksâ&#x20AC;&#x2122; borrowing limits at the Interbank Money Market established within the CBRT were lowered to TL 22 billion by the CBRT. Moreover, it has been lowered to TL 11 billion as of 13 January 2017. It was stated in the announcement made on January 13, 2017 that on the days deemed necessary, the amount of funding provided by the CBRT through Borsa Istanbul repo markets may be limited. Banks will be able to meet their remaining liquidity needs without limits at late liquidity window funding rate at the end of the day. 1

Financial Stability Report- May 2017

Central Bank of the Republic of Turkey Türkiyr a positive effect on deposit growth. It is predicted that the difference between the TL and FX L/D ratios will decrease slightly as a result of the decline in volatility in foreign exchange rates due to

The flat course in L/D ratio has continued since 2015. Chart III.2.4 Loan/Deposit Ratio (Percent, 4-Week Moving Average)

the CBRT's monetary policy in the last few months and the depositors'

140

revived interest in TL deposits.

130

Total

120

While the definition of core liabilities is still a controversial issue, it is generally accepted that deposits are a stable source of funds for banks. In addition to deposits, equities, subordinated debts, long-

04.17

10.16

04.16

10.15

04.15

10.14

04.14

10.13

does not consider non-deposit sources a stable source of funds.

04.13

10.12

account the maturity matching between assets and liabilities and

04.12

10.11

provide a comprehensive measure because it does not take into

10.10

100

following the long-term liquidity positions of banks, it does not

04.11

110

Although the L/D ratio is an important indicator in terms of

Source: CBRT (Latest Data: 05.05.17)

Banks are able to sustain credit growth without weakening the quality of their funding by using relatively stable funding sources.

term issuances and debt items with a maturity of longer than one

Chart III.2.5

year are also sources of funding that can be considered stable for

(Percent)

Loan/(Deposit+Other Stable Sources) Ratio L/(D+Other)*

L/(D+Other)**

banks. In this context, the ratio of banking sector's L/(D + other stable

100

resources) calculated by including the mentioned resources was 82

percent as of March 2017 (Chart III.2.5). This suggests that banks can

100

may be a stable source of funding, provided that the maturity

03.17

09.16

03.16

09.15

03.15

accord, it is estimated that long-term borrowings other than deposits

09.14

Similarly, according to the perspective set out in the Basel III

03.14

09.13

03.13

09.12

03.12

sources by keeping liquidity positions in a safe zone.

09.11

03.11

09.10

meet credit demands with relatively stable long-term funding

Note: (*) Other includes equity, long-term issues, subordinated loans and other loans with maturities longer than one year. (**) Other includes equity, long-term issues, subordinated loans.

Source: CBRT (Latest Data: 03.17)

structure of them is compatible with the assets. In this framework, the Net Stable Funding Ratio (NSFR) has been developed by the Basel Committee in order to measure the long-term liquidity position of banks more extensively and to limit the risks arising from the maturity difference between banks' assets and liabilities. According to the

As a result of the weakening in banks’ demand for foreign resources, there is a limited decline in external debt of the sector. Chart III.2.6 Amount and Growth Rate of Banks’ External Liabilities (Annual Percentage Change, Billion USD)

NSFR, which is expected to be put into practice in Turkey as of 2018, banks will be able to sustain credit growth without weakening the quality of funding by extending the maturities of their foreign debts.1

Growth (Adjusted for FX and Parity Effects) Growth (in USD Term) Amount (RHA)

39 33

180 160

27 140

21 15

-3

In order to monitor the long-term liquidity position of the banks in a more comprehensive manner, the Basel Committee has applied the NSFR taking account of the maturity matching of the assets and liabilities. Details of this ratio are given in Box III.2.1. 1

Financial Stability Report - May 2017

03.17

09.16

03.16

09.15

03.15

09.14

03.14

09.13

03.13

60 09.12

-9 03.11

sector observed since the beginning of 2015 is caused by the

100

03.12

use of foreign resources. The decline in the external debt of the

120

09.11

There has been a limited weakening in the banking sector's

Note: The series that is adjusted for FX and Parity Effects is calculated based on the USD/TRY and EUR/USD parity at end-2013.

Source: CBRT, MKK (Latest Data: 03.17)

Central Bank of the Republic of Turkey Türkiyr

Chart III.2.7 Cost of Syndicated Loans with a Maturity of 367 days (Transaction Based, Percent) Spread (USD) Libor USD CRA Decisions

weakening in banks’ demand for foreign sources, rather than the

Spread (Euro) Euribor Euro

conditions and costs of external borrowing of banks. The weak

0.5

the first quarter of 2017 as well as the increasing risk appetite and

0.0

-0.5

07.14

04.14

03.14

03.17

investments in the recent period and the high credit growth rate in

0.5

10.16

sources. It is estimated that the recovery in the real sector

1.0

08.16

1.5

1.0

05.16

1.5

04.16

investments are believed to have limited banks' demand for foreign

10.15

2.0

08.15

2.0

04.15

2.5

03.15

course in domestic FX credit growth and the slowdown in real sector

2.5

09.14

3.0

fund flows towards the emerging countries will increase the external borrowing of banks. (Chart III.2.6).

Note: Includes only large scale banks. CRA Decisions represent the date of credit rating agencies’ decisions.

Source: PDP (Latest Data: 04.16)

There was a limited increase in the cost of roll-over of external debt. Fluctuations in the borrowing costs of syndicated loans before

The maturities for the banking sector’s external liabilities continue to lengthen. Chart III.2.8

Short Term Medium and Long Term Total (RHA)

120 110 100

Moody's credit rating decision on 23 September 2016 and Fitch's 180

decision on 27 January 2017, there was a limited increase in the

175

borrowing cost margins. The increase in costs is thought to be in line

170

with the additional cost of capital that the creditor institutions will

165

incur depending on the increase in risk weights applied to

160

receivables from banks.

90 80 70 60 50

movements in the Libor and the Euribor interest rates only, as the cost margins remained unchanged (Chart III.2.7). However, after

Change in Banks’ Short Term and Medium-Long Term External Liabilities (Billion USD) 130

the Moody's credit downgrade decision were a reflection of the

155

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

150

12.13

The fact that the increases in the costs of banks have been very limited in the period following the decisions of the credit rating

Source: CBRT, MKK (Latest Data: 03.17)

agencies means that foreign financial institutions have not changed their credit supply. This supports the argument that the deceleration of external borrowing since the beginning of 2015 is largely due to

Chart III.2.9

the preferences of domestic banks rather than foreign investors'

External Debt Roll-Over Ratio (Percent)

125

Short Term Medium and Long Term (RHA)

120

supply constraints. In the same period, the increasing diversity of the 260 240

115

composition of the lender countries is considered to have limited the

220

risks associated with the creditor concentration and alleviated the

105

200

effects of supply-side factors.1

100

180

160

110

140

120

100 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Note: Roll-over ratios are calculated based on 3-month and 12month moving totals of banks’ borrowings and repayments of total external liabilities including securities issued abroad for short term, and medium and long term, respectively.

Fuelled by the supportive measures, the maturity composition of foreign bank debts continues to change in favor of the long term. With the contribution of the regulations introduced by the CBRT to promote long-term non-core liabilities, banks have significantly

Source: CBRT, MKK (Latest Data: 03.17) The details and possible impacts of increasing diversification in the countries/banks that provide funds to the Turkish banking sector are given in Special Topic IV.1. 1

Financial Stability Report- May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Chart III.2.10

reduced their borrowings with up to one year maturity from abroad

External Debt Roll-Over Ratio and its Average Maturity (Percent, Month)

and increased their medium and long-term resources (Chart III.2.8).

120

The long-term external debt roll-over ratio remained well above 100

115

percent for a long time, as the sector preferred to roll over its short-

110

term debt with long-term resources (Chart III.2.9). The long-term and short-term external debt roll-over ratios are now close to 100 percent

Rollover Ratio Weighted Average Maturity (RHA)

60 58 56 54 52

105 50

100

since the transition from short-term to long-term has largely been

the sector's foreign debts has reached 58 months as of March 2017, with a slight decrease in the last two months (Chart III.2.10). In this period, the roll-over of one-year maturity syndicated loans with

44 90

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

achieved in the recent period. The average weighted maturity of

46 95

Note: The external debt roll-over ratio is calculated based on 6month moving totals of banksâ&#x20AC;&#x2122; borrowings and repayments of total external liabilities including securities issued abroad.

Source: CBRT, MKK (Latest Data: 03.17)

maturities of up to three years, the weakened preference for shortterm repo funding, and the tendency to issue long-term bonds Chart III.2.11 Change in External Borrowing Instruments by the End of 2014 (Billion USD) 20

addition, the liquid asset portfolio of the sector provides room for

-10 -15

maneuver for banks to cover FX liquidity shocks even under the most negative scenarios within the one-year window. As of March 2017, the banking sector has foreign currency debt payments of USD 46

Securitization

banking sector to any possible volatility in international markets. In

0 -5

Issues

external debt in favor of long-term increases the resilience of the

Repo

liquidity shocks. The change in the maturity composition of the

Syndication

The banking sector has sufficient liquidity buffers against

Short Term Medium and Long Term

Deposit

composition in favour of the long-term (Chart III.2.11).

Credits

constitute the fundamental dynamics of the change in the maturity

Source: CBRT, MKK (Latest Data: 03.17)

billion and USD 77 billion in the next six months and one year, respectively. In this framework, the developments in global markets as well as the developments in our country will continue to be of particular interest to banks in terms of the roll-over of external debt at favorable maturities and costs. A significant portion of the banks' liquid assets consists of gold and foreign exchange assets held within

Banking sector has sufficient liquidity buffers providing a one-year window for banks to hedge themselves against liquidity shocks. Chart III.2.12 FX Liquid Assets and FX External Liabilities Due Within 1 Year (Billion USD, Percent) Cash Foreign Banks (Free) ROM Reserves FX Liq. Assets/Ext. FX Liab. Due within 1 Year (RHA)

the framework of ROM; and in the second half of 2016 when capital inflows weakened, both the rise in FX costs and the increase in

exchange rates led to a decline in ROM reserves. As of February

2017, there was an increase in ROM reserves as a result of the

strengthening capital movements and falling exchange rate

volatility. In this framework, liquid assets of banks such as cash, free

accounts at foreign banks and ROM reserves are strong enough to

60 55

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

meet half of the foreign debt (Chart III.2.12). The ROM provided by

the CBRT and the FX borrowing facilities with a limit of USD 50 billion

Note: Selected FX Liquid Assets: Cash+Foreign Banks (free) + Required Reserves held within the ROM facility. The dashed line represents 3-month moving average of the FX Liquid Assets / External FX Liabilities Due within 1 Year ratio.

are sufficient to respond to the most negative shocks (Chart III.2.13).

Source: CBRT, MKK (Latest Data: 03.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr In addition to liquid assets, the free Eurobond portfolio of USD 7 billion Chart III.2.13 ROM Reserves + FX Borrowing Facility and External FX Liabilities Due Within 1 Year (Billion USD)

increases the capacity of banks to meet short-term foreign debts.

External FX Liabilities Due within 1 Year

100

ROM Reserves+FX Borrowing Facility from the CBRT

The revival experienced since the beginning of 2016 in FX-

denominated securities issued abroad by the banking sector, which

is highly sensitive to global liquidity developments, continues. The increased risk appetite in the global markets and the monetary

policies of advanced countries’ central banks supporting the 60

liquidity conditions have led to a 13-percent increase in the sector’s 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Source: CBRT, MKK (Latest Data: 03.17)

FX-denominated securities issues abroad since the beginning of 2016. The decrease in short-term issuances and the recent longerterm issuances have resulted in a 70-month average maturity for FXdenominated securities issues abroad (Chart III.2.14). Although the share of FX-denominated securities issued abroad by banks in total

As a result of global liquidity developments, there is a recovery in FX-denominated securities issued abroad by the banks.

FX Issues Abroad

limited since the maturity of foreign securities, which are highly

(Billion USD, Month)

level of 16 percent since the end of 2014. Therefore, the risks related to the FX-denominated securities issued abroad by banks remain

Chart III.2.14

Amount (Stock)

external debts has increased steadily since 2010, it has settled at a

Maturity (RHA)

sensitive to global liquidity developments, is longer than other

borrowing types and continues to lengthen, and since their share in

total external debts is low and stable. In addition, depending on

global liquidity and risk appetite developments, asset backed

securities issues as well as standard securities issues are expected to

26 24 22

20 18 16

56 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

increase in the coming period. The flat course in the amount of the domestic securities issues of the banking sector and in the average maturity of the new issues continues in this financial report period (Chart III.2.15).

Source: MKK (Latest Data: 03.17)

Chart III.2.15 Domestic TL Security Issues (Billion TL, Month)

Amount (Flow) Maturity (RHA)

Amount (Stock)

14 13

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Source: MKK (Latest Data: 03.17)

Financial Stability Report- May 2017

Central Bank of the Republic of Turkey

Box III.2.I

Net Stable Funding Ratio

The global financial crisis showed the importance of developing and monitoring quantitative liquidity standards focusing on long-term stability of the asset and liability structure. Indicators such as loan to deposit ratio that are widely used in measuring funding risks and monitoring liquidity positions of banks fall short of providing an extensive risk assessment as they neglect stable funding sources other than deposits and ignore the maturity composition of assets-liabilities. In this respect, after the global financial crisis, the Basel Committee on Banking Supervision (Basel Committee) published principles for sound liquidity risk management for banks, and developed standards for minimum liquidity ratios. The first among them is the Liquidity Coverage Ratio (LCR) 1 that is intended to promote banks’ resilience against short-term liquidity shocks. The Basel Committee member states have been implementing the LCR gradually since 2015. The second standard is the Net Stable Funding Ratio (NSFR) developed against long-term liquidity risk.

The NSFR aims to limit overreliance on short-term wholesale funding by encouraging a long-term and deposit-based funding structure. According to this ratio, the amount of available stable funding of a bank should be greater than the amount of required stable funding. Unlike the LCR, no gradual transition is foreseen for the NSFR. Therefore, banks have to meet this minimum ratio as of 1 January 2018.

NSFR =

Available Amount of Stable Funding ≥ 100 % Required Amount of stable Funding

The “available amount of stable funding” of a bank is calculated by multiplying the bank’s capital and liabilitiesby the available stable funding (ASF) factors taking into account the residual maturity and counterparty of its liabilities. Basically, regulatory capital, liabilities with effective residual maturity of one year or more, and deposits of retail and SME customers are regarded as stable funding sources while other liabilities are treated with pre-determined discount rates (ASF factors). The “required amount of stable funding” is calculated by classifying assets from the least to the most liquid according to their liquidity-generating capacity and multiplying them by certain required stable funding factors. In this calibration, it is assumed that banks will not roll-over part of their shortterm assets and they will generate funding by encumbering liquid assets or selling them on the market. In this regard, while a required stable funding is sought for assets with a maturity of one year or more, this requirement is less for liquid assets. 2

LCR is calculated by dividing stock of high quality liquid assets to total net cash outflows over the next 30 calendar days. The minimum requirement for total LCR was set at 60 percent in 2015 and it will reach 100 percent by 2019 with a 10 points increase annually. 1

See for funding factors. CBRT Blog, New Era in Liquidity Management in Banking Sector: Net Stable Funding Ratio.

https://tcmbblog.org/en/a-new-era-in-liquidity-management-in-banking-sector-net-stable-funding-ratio/

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Figure III.2.I.1

Assets

Liabilities and Equity

('Use of Funds')

('Source of Funds')

Cash and Central Bank

Regulatory Capital

High Quality Liquid Assets

Long Term Liabilities

Short Term Loans to Financial Institutions Short Term Retail and Corporate Loans Residential Mortgages

Retail Deposits Commercial Deposits Wholesale Funding

Long Term Loans

Stability of Liabilities Increases

Liquidity of Assets Decreases (Stable Funding Requirement Increases)

Ranking Assets by Liquidity and Liabilities by Stability

Other Liabilities Other Assets

Although the NSFR reporting has not yet started in the Turkish banking sector, the NSFR of the sector can be calculated under certain assumptions. The strongest assumption made within the scope of the study is to calculate the NSFR based on original maturities.3 In fact, all funding and asset items should be considered in view of the residual maturity in the NSFR calculation. However, the fact that the maturity of deposits that account for a significant portion of the funding is shorter than six months provides reasonable grounds for our assumption on the funding side. Whereas, on the asset side, although the fact that original maturities are longer than residual maturities poses a risk of underestimating the NSFR, it supports a cautious stance. The NSFR that is calculated for the Turkish banking sector is above 100 percent which is the minimum ratio set in the Basel standard.4 In this framework, it is possible to ascertain that banks finance their assets with stable funding sources.

In NSFR calculation, assets and liabilities are considered in view of the original maturity. Commercial deposits below TL 250 thousand are considered as SME deposits. All encumbered securities and the collateral for reverse repo receivables are deemed as â&#x20AC;&#x153;level 1â&#x20AC;? assets. The initial margin considered as 85 percent in derivative transactions is not decomposed and taken as 100 percent among other on-balance sheet items. Net receivables from derivative transactions are taken into account along with the required funding factor of 100 percent, regardless of the collaterals of transactions. 3

The Basel Committee conducts a semiannual quantitative impact study to evaluate capital and liquidity ratios within the scope of Basel III. These studies are based on the data of a total of 230 reporting banks. Three banks from Turkey also contribute to this reporting. The weighted average NSFRs for group-1 and group-2 banks by June 2016 were 114 percent and 115 percent, respectively. Banks that contributed to reporting from Turkey meets the minimum NSFR as of this date and their weighted average NSFR at 114.6 percent remains above the average of group-1 banks in which they are listed. 4

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Chart III.2.I.1 Net Stable Funding Ratio in the Turkish Banking Sector (Percent)

110 NSFR

Basel Regulatory Minimum

100

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

Source: CBRT

To conclude, the NSFR, which will be effective in 2018, will enable comprehensive measurement of the liquidity risk by taking into account the maturity of the assets and liabilities on banksâ&#x20AC;&#x2122; balance sheets. From a long-term perspective, the NSFR will be a supplementary liquidity ratio to the LCR that measures the resilience of the banking sector against short-term liquidity shocks. In light of current measurements, the long-term liquidity positions of Turkish banks are strong. Adopting the NSFR as another minimum requirement for banks in addition to the LCR would be important for the sustainability of this outlook.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

III.3 Interest Rate and Exchange Rate Risk Chart III.3.1 Short Term Open Position in TL (Ratio to Total TL Liabilities, Percent)

Sudden changes in interest rates affect banks primarily via two

channels: the repricing channel, since maturity structures of assets

and liabilities differ, and the balance sheet channel since the value of

0 -5

securities in their trading portfolio changes as interest rates change. In

-10

this respect, analyzing the maturity structure of assets and liabilities

-15

-20

and the characteristics of securities portfolio is important to make a

-25 -30 03.13 0-1 Month

03.14

03.15

1-3 Months

03.16 3-6 Months

09.16

03.17

6-12 Months

sound evaluation of the sensitivity of the sector to interest rate fluctuations.

Source: CBRT (Latest Data: 03.17)

Weighted with repricing periods, the average maturity of interest rate-sensitive TL assets is 20 months and that of interest rateTransition to FX deposits increases open position in shorter maturities.

sensitive TL liabilities is 3 months. Recently, both maturities have shown signs of extension. On, the FX side, the average maturity of assets is about 21 months and that of liabilities is approximately 12 months. The

Chart III.3.2 Short Term Open Position in FX

outlook of interest rate-sensitive TL and FX positions are similar. The

(Ratio to Total FX Liabilities, Percent)

open positions, which are relatively higher in shorter maturities,

6 4

decrease as maturities extend and turn into long positions. On the FX

side, the ratio of open positions with a maturity of up to one month to

0 -2

the total interest rate-sensitive liabilities has recently increased.

-4 -6

Basically, this rise stemmed from the increasing demand for FX

-8 -10

deposits, and it increases the potential risks posed by the upward

-12

-14 03.13 0-1 Month

03.14

03.15

1-3 Months

03.16 3-6 Months

09.16

03.17

6-12 Months

movements in foreign interest rates on banksâ&#x20AC;&#x2122; income statements (Chart III.3.1 and Chart III.3.2).

Source: CBRT (Latest Data: 03.17)

In light of these premises, balance sheet and off-balance sheet Grafik III.3.3 Interest Rate Risk via Repricing Channel Measured with Economic Value Approach TL 0

FX (RHA ve Upper A.)

0.2 0.4 0.6 0.8

1.2 1.4 1.6 1.8

Loss/Capital (Percent)

-2

-6

hikes with a magnitude of 5 points for TL and 2 points for FX. In this respect, the potential loss was estimated by using the economic value

-0.2

-4

positions of the banking sector have been exposed to interest rate

approach and its ratio to capital was inspected. Accordingly, a TL shock of up to 5 points is estimated to generate a loss up to 16 percent

-0.4

-8

-0.6

-10 -12

-0.8

-14

of capital whereas an FX shock up to 2 points is estimated to generate a more limited impact, up to 1 percent of capital.

-1

-16

-18

-1.2 0

0.5

1.5

2.5

3.5

Interest Rate Shock (Points) Source: CBRT, Authorsâ&#x20AC;&#x2122; own estimation (Latest Data: 03.17)

4.5

Other than repricing, another channel that may affect financial intermediation industry is the securities revaluation channel. Although the share of securities in total assets has been declining, the impact of the changes in values of securities in trading portfolio driven

Financial Stability Report - May 2017

Türkiye Cumhuriyet Merkez Bankası Türkiyr by interest rate variations on capital is still important. This effect has been tested for an interest rate hike up to 5 points on the TL side and

Chart III.3.4 Interest Rate Risk on Securities with Fixed Interest Rate in Trading Portfolio

2 points on the FX size across all maturities. Accordingly, the 0.0

estimated to be up to 3 percent and up to 2 percent of capital for TL

-0.5

and FX, respectively (Chart III.3.4).

The Turkish banking sector preserves its resilience against direct

Loss/Capital (Percent)

prospective losses in capital via securities revaluation channel are

TL 0.2 0.4 0.6 0.8

0.5

FX (Upper A.) 1 1.2 1.4 1.6 1.8

-1.0 -1.5 -2.0 -2.5

-3.0

FX risks via balance sheet items. It is observed that banks are holding

-3.5

their open positions on their balance sheets at reasonable levels and they are quite prudent in hedging these positions with off-balance sheet transactions. As a result, the FX net general position

1 1.5 2 2.5 3 3.5 4 Interest Rate Shock (Points)

4.5

Source: CBRT, Authors’ own estimation (Latest Data: 03.17)

(FXNG)/Capital ratio is close to zero level, well below the two-sided legal threshold of 20 percent (Chart III.3.5).

FX net general position is close to zero level. Chart III.3.5 FX position in the Banking Sector

An analysis of the off-balance sheet FX transaction items by

(Billion USD, Percent)

types reveals that currency swaps are the primary instrument in FX risks management and their weight has increased over time. Meanwhile, short and long-dated FX transaction commitments along with FX

3 25

2 1

derivative options are preferred less in managing FX position. Nevertheless, the magnitude of currency swap transactions increases

-1

-25

-2 -3

the sensitivity of profitability to currency swap rates (Chart III.3.6).

-4 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

-50

In-Balance Sheet Position Off-Balance Sheet Position FXNG/Capital (RHA)

Currency swaps, increasing their weights, preserves their importance in FX position management.

Source: CBRT (Latest Data: 03.17)

Chart III.3.6 Shares of Gross Positions of Off-Balance Sheet FX Transaction (Assets + Liabilities) Currency Swaps

FX Transactions

Options

Other Derivatives

100% 80%

60% 40% 20%

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

03.13

12.12

Source: CBRT (Lates Data: 03.17)

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

III.4 Profitability and Capital Adequacy

Profitability continues to recover. Chart III.4.1 Return on Assets and Return on Equities (ROE) (Percent)

The banking sector profitability indicators, which have been

Dispersion of ROA (10 Largest Banks) Return on Assets (Sector) Return on Equities (Sector) (RHA)

trending upwards since the last quarter of 2015, continued to rise in 2017 despite a flattening trend in the final quarter of 2016. While the

2.5

profitability performance, the profitability indicators of large banks in

1.5

the sector have converged (Chart III.4.1). While the capital

adequacy ratios (CAR) were pulled down due to the rapid rise in risk-

0.5

weighted assets because of the recovery in TL loans and

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

recent movement is more evident in banks with relatively low

Note: Profitability ratios are calculated by dividing the annual cumulative profit by one year's average denominator.

appreciation effects in FX assets; these ratios have attained their levels recorded in the third quarter of 2016 on the back of the slightly

Source: CBRT (Latest Data: 03.17)

decreased capital adequacy ratios (CAR), with the stabilization in FX rates, increase in profitability and regulation introduced for risk

Capital adequacy preserves its vigorous position.

weights (Chart III.4.2).

Chart III.4.2 CAR and Core Tier 1 CAR

III.4.1

(Percent)

Developments in Profitability

Dispersion of CAR (10 Largest Banks)

CAR (Sector) Core Tier 1 Car (Sector)

An analysis of the factors affecting return on assets (ROA) based on income statement items reveals that: the improvement in net interest income, the perpetuation of austerity measures in non-

interest expenses and the trend of securities, foreign exchange and 14

derivatives position have been positive whereas the impact of the limited increase in NPLs on collaterals has been negative. Tax 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

provisions are based on net income and the higher the gross profit, the lower become the net profits (Chart III.4.3).

Source: CBRT (Latest Data: 03.17)

Chart III.4.3

Over the last year, the impact of net interest income on ROA

Effects of Income Statement Items on ROA (Points)

1.8

has been around 22 basis points. In this period, the impact from

0.25

widening interest rate spreads on net interest income has become 0.11

1.6

stronger. The rate cuts of the CBRT last year contributed to this

0.16

1.64

0.22

1.4

0.09

costs and the increased probability of the Fed continuing to raise

0.02

1.2

rates is expected to limit the favorable impact coming via this

1.19

channel. On the other hand, the revival of loans with the increase in

Source: CBRT (Latest Data: 03.17)

ROA 03.2017

Tax Provisions

Securities, FX, Derivatives

Non-Interest Expenses

Non-Interest Income

NPL Provisions

Net Interest Income

ROA 03.2016

phenomenon as well. However, the recent reversal of TL funding

the economic activity and the easing in rigidity in financial conditions with the help of the KGF stimulus are expected to play a positive part in net interest income via the volume channel (Chart III.4.4).

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

Over the last 12 months, the uptrend in NPL ratios decreased ROA around 9 basis points through provisions. Nevertheless, this

Chart III.4.4 Contribution to Changes in the Net Interest Income (Annualized Billion TL, Percent)

IR Margin Effect Volume Effect Net Interest Income Net Interest Margin (RHA)

uptrend has been partly decreased with the contribution of the recent loan expansion. For all that, the increase in NPL provision rates indicates that the banking industry preserves its prudential stance against probable risks. On the other hand, while the fading effects of refunds of fee and commissions collected in the past years

4.8 6 4.6 0 4.4

considerably limit the adverse impact of these items on profitability.

profitability as the austerity measures on operational expenses continue. In the forthcoming period, the non-interest income outlook is expected to be partly limited as the one-off inflows during 2016 wane.

-6

4.2

-12

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

Moreover, non-interest expenses are contributing to increasing

5.2

Source: CBRT (Latest Data: 03.17)

Chart III.4.5 Profit/Losses from Security, Derivative and FX Transactions (Annualized Billion TL)

The outlook of the other non-interest income/expenses item, in

and exchange transactions, turned positive over the last one-year

period. Although the profit derived from securities trading was

marginal compared to previous periods, the decline in the net swap

-5

position due to the slowdown in loan expansion and the decrease in

-10

swap rates last year is believed to play a part in this positive outlook.

-15

Nevertheless, the increasing demand for swaps with the recent recovery in loan growth and the reversal in interest rates are

from FX Transactions

from Derivatives

Total Profit/Loss

12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

which banks recognize their position in securities trading, derivatives

from Securities

Source: CBRT (Latest Data: 03.17)

expected to restrict this positive outlook in the upcoming period (Chart III.4.5).

Chart III.4.6 Changes in Items Affecting Capital (Annualized Billion TL)

III.4.2

Capital Adequacy 60

The significant rise in profitability was the main driver of the increase in the legal capital over the last year. In this period,

50 40 30

because of the upward movement in government bond rates, the

equity capital was adversely affected via the securities revaluation

channel and this negative impact contained the strong positive

new acquisitions of subordinated debts started to offset the adverse

Financial Stability Report - May 2017

-10 12.12 03.13 06.13 09.13 12.13 03.14 06.14 09.14 12.14 03.15 06.15 09.15 12.15 03.16 06.16 09.16 12.16 03.17

impact from the fixed assets revaluation channel. Meanwhile, the

Adjustment and Evaluation Items Profit, Reserves and Paid Capital Balance Sheet Capital Subordinated Debts Regulatory Capital Other Items

Source: CBRT (Latest Data: 03.17)

Central Bank of the Republic of Turkey Türkiyr impact of previously acquired subordinated debt excluded from the

Chart III.4.7 RWA Components

legal capital definition by the regulations1 introduced (Chart III.4.6).

(Annualized Billion TL)

100

100 70

There has been no significant change in the risk-weighted

asset (RWA) composition over the last year, and credit risk preserved

10 90

its dominant position with a 90 percent share. Credit risk, which had been increasing until recently due to the depreciation of TL and the

-20

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

-50

12.13

recovery in TL loans, started to trend downward on the back of the rebalancing of exchange rates and the BRSA’s lowering of the risk weights of FX required reserves to zero. The downtrend in credit risk is

Operational Risk (Share) Market Risk (Share) Credit Risk (Share) Operational Risk (Growth, RHA) Market Risk (Growth, RHA) Credit Risk (Growth, RHA)

expected to continue as the volume of KGF guaranteed loans expand. The operational risk amount, which is generally updated once a year by banks, increased by about 16 percent year-on-year

Source: CBRT (Latest Data: 03.17)

(Chart III.4.7). Chart III.4.8 CARs According to Bank Types (Percent) Public Deposit Domestic Private Deposit Foreign Deposit Participation Development and Investment (RHA)

18 17

credit growth rate that was lower compared to previous periods has 36

adequacy of the sector grew stronger over the last year. This impact

is evident in all banking groups except development and investment

14 13

banks, which have already high CARs, and domestic private banks

because of the exclusion of some banks from this category due to

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

09.14

06.14

03.14

12.13

09.13

06.13

03.13

restricted the growth in RWA. As a result of these effects, the capital

32 15

12.12

While the rise in profitability has strengthened capital, the

changes in ownership (Chart III.4.8).

Source: CBRT (Latest Data: 03.17)

The relationship between CAR and loan growth is analyzed utilizing the method and parameters used in Box III.4.1 of the 2015 Chart III.4.9

November Financial Stability Report taking into account the latest

Banking Sector ‘s Capacity of Supporting Loan Growth under Current Profitability and Capital Adequacy Levels 18

profitability levels. Accordingly, if the banking sector preserves the

CAR (10.5 Percent)

growth rate of 15 percent without decreasing the current CAR level.

CAR (12 Percent)

14 Number of Years

current ROA level of 1.64 percent, it would be able to support a loan

With such a profitability level, a loan growth rate of 20 percent can

12 10

be supported for up to six years and a loan growth rate of 30

percent can be supported for up to three years without going under

6 4

the target CAR level (12 percent). Finally, if the CAR limit is assumed

2 0 16

Loan Growth Rate (Percent)

Source: CBRT (Latest Data: 03.17)

to be the legal minimum limit including capital preservation buffer (10.5 percent), the industry will not face any restriction of capital adequacy while supporting loan growth under 19 percent (Char III.4.9). Regulation on the Amendment to the Regulation on Equities of Banks (O.G. No.29511 of 3.10.2015 and O.G. No.29599 of 20.01.2016) 1

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey

Box

Exchange Rate Developments and The Capital Adequacy Ratio

III.4.I

Banks, like other companies, obtain funding in two different ways: borrowing and equity. As a core source of funding, equity has no maturity and does not create a repayment obligation. In addition, equity also acts as a buffer to cover losses in periods when banks incur losses. Since equity has a risk-free structure in the funding of banks, regulatory authorities require banks to meet at least some portion of their funding need via equity. This ratio, which is known as "Basel Capital Adequacy" in the regulatory literature, is calculated by dividing the regulatory capital by risk weighted assets:

đ??śđ??´đ?&#x2018;&#x2026; =

đ?&#x2018;&#x2026;đ?&#x2018;&#x2019;đ?&#x2018;&#x201D;đ?&#x2018;˘đ?&#x2018;&#x2122;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;Ś đ??śđ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;&#x2026;đ?&#x2018;&#x2013;đ?&#x2018; đ?&#x2018;&#x2DC; đ?&#x2018;&#x160;đ?&#x2018;&#x2019;đ?&#x2018;&#x2013;đ?&#x2018;&#x201D;â&#x201E;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2019;đ?&#x2018;&#x2018; đ??´đ?&#x2018; đ?&#x2018; đ?&#x2018;&#x2019;đ?&#x2018;Ąđ?&#x2018;

The capital adequacy ratio (CAR) is affected by changes in the regulatory capital and risk weighted assets. The regulatory capital increases with the positive contribution of new capital inflows and profitability. Since almost all of the subordinated debt is denominated in foreign currencies and is included in the regulatory capital, exchange rate hikes lead to an increase in the regulatory capital. On the other hand, that increase is limited since the share of subordinated debts in total regulatory capital is low. Risk weighted assets rise due to increases in average risk weights and the acceleration in total asset growth. The CAR is affected negatively by the increase in risk weighted assets and positively by the decrease in risk weighted assets. The risk-weighted assets of banks increase as the TL equivalent of the total assets of the banking sector increases during periods of escalating exchange rate. Therefore, in periods of TL depreciation, the CAR moves downward. In this study, the effect of exchange rate movements on CAR is analyzed using a recent case study. The analysis was carried out using the end-October and end-November periods of 2016, when the value of the basket rose by 9.1% against the TL. In this period, the CAR of the banking sector decreased from 16 to 15.3 percent. In addition to the increasing effect of exchange rate on foreign currency denominated assets, the growth of other TL assets and the portfolio preferences were also effective in this decline. The exchange rate-driven increase in the regulatory capital did contribute positively to the CAR. In Table III.4.I.1, however, only the exchange rate-driven negative impact on the CAR is calculated through decomposition of the effects mentioned above.

The exchange rate adjustment is based on the assumption that at the end of November 2016, the exchange rate maintains its level at the end of October. On the other hand, the limited positive effects of the increase in the exchange rate on non-FX regulatory capital and TL assets are ignored. According to the calculations, the banking sector CAR moved downward by 41 basis points as a result of the 9.1 percent increase in the exchange rate basket in the October-November period. The changes in non-FX regulatory capital and in TL assets as well as the portfolio preferences have a downside effect on the CAR of about 27 basis points (Chart III.4.I.1). It can be concluded that with a linear calculation, a 10-percent depreciation of the currency causes the CAR to decrease by about 45 basis points in the short term.

Financial Stability Report May 2017

Central Bank of the Republic of Turkey

Table III.4.I.1 Exchange Rate Effect on CAR Calculation (Unless Otherwise Stated, Billion TL)

October 2016 3.19

November 2016 3.48

TL Assets

1,575

1,607

FX Assets

1,007

1,095

Total Assets

2,581

2,702

Basket Exchange Rate (TL)*

FX Assets (billion FX basket)

315

314

1,003

Exchange Rate Adjusted Total Assets (a)

2,610

Risk Weighted Assets/Total Assets (%) (b)

82.3

82.5

Total Regulatory Capital (c)

339

340

FX Regulatory Capital (d)

Exchange Rate Adjusted Assets**

Exchange Rate Adjusted Regulatory Capital*** (e)

Exchange Rate Adjusted Total Regulatory Capital (c-d+e)

338

2,123

2,230

2,154

16.0

15.3

Exchange Rate Adjusted Capital Adequacy Ratio**** (%)

15.7

Exchange Rate Effect (%)

0.41

Total Risk Weighted Assets (f) Exchange Rate Adjusted Risk Weighted Assets (a*b) Capital Adequacy Ratio (%) (c/f)

* Composed of 70 percent USD and 30 percent Euro. ** Obtained by multiplying FX basket denominated value of end-November 2016 FX assets by the end of October FX basket value. *** Subordinated debt assumed to constitute all FX regulatory capital and obtained by multiplying FX basket denominated value of end-November FX regulatory capital by the end of October FX basket value, **** Calculated by dividing exchange rate adjusted total regulatory capital by exchange rate adjusted risk weighted assets.

Source: BRSA, CBRT calculations

Chart III.4.I.1 Decomposition of the Effects on CAR (Percent)

15.95

0.41

0.27

15.27

10 October CAR

Exchange Rate Effect

Other Effects

November CAR

Source: BRSA

It should be underlined that the calculations made are based on a partial analysis and do not reflect dynamic effects. For example, the increase in exchange rate has a downward effect on the short-term CAR, while in the md-term some of these effects have the potential to narrow on the back of an increase in profitability. Since the income from FX assets is in the form of FX, the profitability will also be positively affected from an increase in the exchange rate. Given that profitability is one of the major sources of equity growth, the CAR, which moves downward in the short run due to exchange rate developments, may rise in line with the increase in profitability in the longer run.

Financial Stability Report May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

IV. Special Topics IV.1

Global Liquidity and Regional Distribution of Cross-Border Bank Loans

Abstract

This study analyzes the impacts of increasing diversification in the countries/banks that provide funds to the Turkish banking sector on the sensitivity of cross-border bank loans to the global liquidity conditions. Recently, we observe that the Turkish banking sector has not only continued to roll over its debts from traditional financial centers but also gained access to new countries and banks. As a result, banks based in non-traditional financial centers have become an important source of funding for the Turkish banking sector. Moreover, our estimation results indicate that the increasing diversification of lender origin lowers the sensitivity of cross-border bank loans to global liquidity conditions and therefore limits the spillovers of potential financial shocks that may occur in systemically important sources.

IV.1.1

Introduction

Global banks and financial institutions have significantly increased their international activities over the last twenty years. With the rise in international activities, financial integration has also deepened and gained strength globally, and capital flows from advanced countries to emerging countries have increased. The strong growth in cross-border capital flows has added to the importance of external financing structure in terms of risks that may emerge regarding the roll-over of loans in emerging countries. The transmission of financial shocks and country or bank specific events to other countries through these capital flows underlines the importance of the diversification of lender countries or banks for borrowers. In this context, we observe that the Turkish banking sector has not only continued to roll over its debts from traditional financial centers but also gained access to new countries and banks. This special topic shares the findings which suggest that the increasing diversification of lender origin lowers the sensitivity of cross-border

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr bank loans to global liquidity conditions and therefore limits the spillovers of potential financial shocks that may occur in systemically important sources. Capital flows to emerging markets, including Turkey, may be in the forms of direct investment, portfolio investment, bank loans and debt securities. There was strong growth in the capital flows to Emerging Asia, Latin America and Emerging Europe until 2007 and then pronounced contractions in all fund types in these regions during the 2008 crisis. The expansion of foreign direct investment, bank loans, portfolio equity and net debt securities was followed by a steep reversal in all broad categories of inflows, with by far the sharpest decline in international bank loans (Cetorelli and Goldberg, 2011). Similarly, Lane (2014) finds that the international bank loans were affected by the global crisis more than foreign direct investments

and

portfolio

investments.

Therefore,

the

World

Economic Outlook (WEO) report, released by the IMF in April 2009, argues that global bank linkages â&#x20AC;&#x153;fuel the fireâ&#x20AC;? for the spreading of the current crisis to emerging markets. The literature has established that the classical push and pull factors are the main determinants of cross-border bank loans, which are relatively more sensitive to global developments than other types of capital flows. The push factors are related to common external conditions that mobilize loan flows such as the global risk appetite and uncertainty, funding conditions of banks that play an important role in the allocation and intermediation of global liquidity, and monetary and liquidity policies of advanced countries. In particular, the severe impact of the 2008 global financial crisis on global financial conditions has caused significant volatility in capital flows to emerging countries, and has become a risk factor for financial stability. In this context, the transmission of financial shocks in a certain region to other countries through financial linkages led borrowing countries to take measures against the negative effects induced by the volatility in capital flows. In this period, several countries

including

Turkey

started

implementing

series

macroprudential policies to reduce the volatility stemming from the changes in global liquidity conditions and to support financial stability. Along with such policies, Cerutti et al. (2014) study a number of borrower country characteristics, specifically indexes of exchange rate flexibility, capital controls, the overall institutional environment, and

bank

regulation

(the

strength

capital

adequacy

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr requirements, supervisory powers, and limits on foreign bank

Chart IV.1.1

presence) and find that these factors reduce a country’s exposures

Number of Lender Countries1

to changes in global liquidity.

120

(Percent)

Sector Largest 7 Bank*

110

After the global financial crisis, in an environment of

100

macroprudential policies implemented by policy makers in Turkey,

10.16

03.16

08.15

01.15

06.14

11.13

global liquidity conditions and therefore limits spillovers of potential

04.13

09.12

diversification

02.12

increasing

07.11

the

12.10

countries/banks lowers the sensitivity of cross-border bank loans to

that

05.10

evaluate

01.08

lender

10.09

03.09

financial centers and also gained access to new countries/banks.

08.08

the sector has continued to roll over its debts from traditional

financial shocks that may occur in systemically important sources. In

(1) Excludes external debt issuances. (*) Represents the number of lender countries that provide funds to the 7 banks with the largest asset size in the Turkish banking sector

this context, we first test the sensitivity of cross-border bank loans to

Source: CBRT (Latest Data: 01.17)

global liquidity indicators, and then analyze the effects of the increasing diversification of lender countries/banks on this sensitivity. Chart IV.1.2

IV.1.2

Concentration of Lender Countries

Share of Lender Countries with the Highest Amount of Debt in Total Debt1 (Based on Headquarters of Lender Banks, Percent)

border loans, and rolled over its loans from a larger number of

countries/banks (Chart IV.1.1). The increasing diversity of Turkish

banks’ external funding across countries and banks in the recent

63 58

48 43 35

10/16

03/16

08/15

01/15

06/14

11/13

04/13

09/12

02/12

07/11

12/10

05/10

specific only to large banks, but is also observed in relatively small banks in terms of asset size.

10/09

increasingly higher number of sources. Moreover, this situation is not

03/09

changed significantly, implies that banks roll over their debts from an

01/08

period, particularly when the amount of external debts has not

3 Countries 5 Countries 10 Countries

08/08

The Turkish banking sector has steadily increased its cross-

(1) Excludes external debt issuances.

Source: CBRT (Latest Data: 01.17)

Charts IV.1.2 and IV.1.3 show the share of lender countries with the highest amount of debt in total debt of the Turkish banking sector. The increasing diversity of Turkish banks’ external funding across countries and banks was also reflected in the amount of debt; due to the increase in the number of lender countries, the share of countries with the highest amount of debt in total debt decreased steadily.

Chart IV.1.3 Share of Lender Countries with the Highest Amount of Debt in Total Debt1 (Based on the Country of Residence of Lender Banks, Percent)

3 Countries 5 Countries 10 Countries

90 85 80 75

The diversification of Turkish banks’ external funding across

countries and banks also affects the regional distribution of the cross-

border bank loans. Traditionally, it is known that the banks

headquartered in the Eurozone, the USA and the UK are at the

50 42

Turkish banking sector’s external borrowing from regions outside the

(1) Excludes external debt issuances.

Eurozone, the USA and the UK in total debt has been gradually

Source: CBRT (Latest Data: 01.17)

Financial Stability Report - May 2017

10/16

03/16

08/15

01/15

06/14

11/13

04/13

09/12

02/12

07/11

12/10

05/10

10/09

03/09

08/08

01/08

forefront of the global banking network. However, the share of the

Central Bank of the Republic of Turkey TĂźrkiyr increasing in the recent years (Charts IV.1.4 and IV.1.5). Due to the

Chart IV.1.4 Regional Distribution of External Debt 1

debt crisis in the Eurozone in 2011, the share of the Eurozone-based

(Based on Headquarters of Lender Banks, Percent)

Euro Area Britain

US Others

banks, which have an important role in the sector's foreign funding, has visibly decreased over the years. Similarly, the share of the UK-

07.15

01.17

banking sector as it has at the same time gained access to new

07.16

have also become an important source of funds for the Turkish

10 01.16

01.15

07.14

centers. However, banks based in non-traditional financial centers

01.14

07.13

01.13

07.12

the sector continues to roll over its debts from traditional financial

01.12

07.11

01.11

07.10

01.10

(1) Excludes external debt issuances.

based banks has been steadily declining since mid-2014. As a result,

countries/banks. Recently, there have been financial problems in some regions

Source: CBRT (Latest Data: 01.17)

from which the Turkish banks intensively borrow. For instance, in 2016, the profitability and the stock value of the Eurozone-based banks dropped (Chart IV.1.6). In addition, after the Brexit decision,

Chart IV.1.5

developments in the UK-based banks, which have played an

Regional Distribution of External Debt 1 (Based on the Country of Residence of Lender Banks, Percent)

Euro Area Britain

important role in the allocation and intermediation of global liquidity,

US Others

are also closely monitored by international financial markets. The

5 01.17

07.16

01.16

07.15

01.15

07.14

01.14

07.13

affect the intermediary capacity of the aforementioned banks

01.13

07.12

01.12

07.11

persistence of these problems or uncertainties has the potential to

01.11

07.10

01.10

(1) Excludes external debt issuances.

negatively. It is obvious that the Turkish banking sector is not experiencing any problems in rolling over debt from either the Eurozone or the UK banks. Accordingly, it should be emphasized that the increased diversification of Turkish banksâ&#x20AC;&#x2122; external funding across countries and banks in recent years is a favorable development mitigating the risks that may stem from the Eurozone or the UK

Source: CBRT (Latest Data: 01.17)

banking system. In this context, we conduct an empirical analysis to

Chart IV.1.6

understand how the increasing diversification in the external sources

Bank Indexes in Euro Zone and USA1 (Percent)

KBW Bank Index* Stoxx Europe 600 Index Stoxx Europe 600 Banks Index

of cross-border bank loans to global liquidity conditions. First, we test

125

the sensitivity of cross-border bank loans to global liquidity indicators.

115

105

increasing diversification on the sensitivity of cross-border bank loans

to global liquidity indicators.

01.16 01.16 02.16 03.16 03.16 04.16 05.16 05.16 06.16 07.16 08.16 08.16 09.16 10.16 10.16 11.16 12.16 12.16 (1) Indexed to January 2016=100 (2) KBW Bank Index serves as a benchmark of the US banking sector

Source: Bloomberg (Latest Data: 31.12.16)

that provide funds to the Turkish banking sector affects the sensitivity 135

135

We then analyze, using the panel data method, the effects of

IV.1.3

Data Set and Methodology

There are a number of liquidity indicators which the existing theoretical and empirical literature has found relevant in terms of their impact on global fund flows. These indicators are:

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr 

Uncertainty and Risk Aversion: Uncertainty and risk appetite refer to a combination of multiple factors – macroeconomic conditions, lenders’ and borrowers’ risk appetites, and the monetary policy. In the empirical literature, uncertainty is commonly monitored through the S&P 500 VIX Index, the stock option prices-based measure of implied volatility.



Global Banks’ Funding Conditions: The lending appetite and the funding conditions of banks, which play a role in the allocation and intermediation of global liquidity, constitute one of the primary indicators affecting capital flows. Although a number of measures are used for global banks’ funding conditions and lending appetite in the literature, measures such as the TED spread, the real credit growth or the ratio of private credit to GDP are particularly employed.



Monetary Policy: This includes the general level of interest rates and the slope of the yield curve. Although the effect of low interest rates on banks’ risk-taking is supported by some empirical literature, its economic significance and precise causal channels remain the subject of much debate. In contrast, the effect of the term premium on banks’ risk-taking has a clearer economic implication. Banks borrow short-term and lend long-term, so their domestic investment opportunities are less profitable when the yield curve is flat. This may trigger banks’ search for yield, including in the form of cross-border bank loans.



Money Aggregates: The empirical literature also points out that the growth in some components of broad money measures, such as wholesale or non-financial enterprises’ deposits, can complement leverage measures in explaining bank risk as they indicate the relative ease of funding conditions. Table IV.1.1 Global Liquidity Indicators US VIX

CBOE S&P500 Volatility VIX

US TED spread

3-month TED spread (LIBOR - Treasury bill)

US slope of yield curve

10 year/3 month US Treasury yield spread

US real policy rate

Federal Funds Target Rate (deflated with CPI)

US growth rate of real credit

Annual growth rate of real private credit

US credit-to-GDP ratio

Private credit/GDP

US growth rate of M2

ABD M2 Parasal Göstergesindeki Yıllık Büyüme Oranı

In this context, we use the following model including the fixed effects panel data method to analyze the effects of global liquidity

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr indicators, summarized in Table IV.1.1, on the Turkish banksâ&#x20AC;&#x2122; crossborder loans: 1 Li,c,f,t = Î˛0 + Î˛1 (Global)tâ&#x2C6;&#x2019;1 +Bank â&#x20AC;˛i,tâ&#x2C6;&#x2019;1 Îą + TRâ&#x20AC;˛tâ&#x2C6;&#x2019;1 Î´ + Îłi +Âľc +Éłf + Î¸t +Îľi,c,f,t where the dependent variable Li,c,f,t is the logarithmic value of bank iâ&#x20AC;&#x2122;s cross-border loans obtained from country c in type f at time t; (Global)tâ&#x2C6;&#x2019;1 is the value of global liquidity indicators at time t-1; Bank â&#x20AC;˛i,tâ&#x2C6;&#x2019;1 is the balance sheet ratios of bank i at time t-1; TRâ&#x20AC;˛tâ&#x2C6;&#x2019;1 is the value of macro indicators of the Turkish economy at time t-1; Îłi is bank iâ&#x20AC;&#x2122;s fixed effects, Âľc is lender country câ&#x20AC;&#x2122;s fixed effects, Éłf is loan type fâ&#x20AC;&#x2122;s fixed effects, Î¸t is the fixed effects for time t. The estimation results obtained by the said model are given in Table IV.1.2. Then, in order to test the effect of the increasing diversification on the sensitivity of cross-border bank loans to global financial cycles, we add the interaction between global liquidity indicators and the number of lender countries that provide external funds to Turkish banks. Li,c,f,t = Î˛0 + Î˛1 (Global)tâ&#x2C6;&#x2019;1 + Î˛2 (Global â&#x2C6;&#x2014; Country)tâ&#x2C6;&#x2019;1 â&#x20AC;˛ â&#x20AC;˛ + Î˛3 (Country)tâ&#x2C6;&#x2019;1 + đ??ľđ?&#x2018;&#x17D;đ?&#x2018;&#x203A;đ?&#x2018;&#x2DC;đ?&#x2018;&#x2013;,tâ&#x2C6;&#x2019;1 Îą + đ?&#x2018;&#x2021;đ?&#x2018;&#x2026;tâ&#x2C6;&#x2019;1 Î´ + Îłi + Âľc+ Éłđ?&#x2018;&#x201C; + Î¸t + Îľi,c,f,t

(Country)i, tâ&#x2C6;&#x2019;1 represents the number of lender countries that provide external funds to bank i at time t-1. The estimation results obtained by the said model are given in Table IV.1.3.

IV.1.4

Empirical Findings

In Table IV.1.2, we test the sensitivity of the Turkish banking sectorâ&#x20AC;&#x2122;s cross-border loans to global liquidity conditions. In order to control the demand side, we add the balance sheet ratios of the borrowing banks and the macro indicators of the Turkish economy to the model.2 As also identified in the existing theoretical and empirical literature, the US global liquidity factors are statistically significant drivers of cross-border bank flows when considered individually. VIX and TED spreads have the expected negative signs, indicating that cross-border

flows decrease

during times of

uncertainty. The real credit growth and the ratio of private credit to Analysis is based on the monthly external debt data of Turkish banks for the period between December 2002 and December 2016. We add the lagged values of the explanatory variables to the regression in order to eliminate the possible endogeneity problem. 1

Time-invariant and unobservable factors related to the borrower bank, lender country and loan type have also been controlled via fixed effects. 2

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr GDP in the US have the expected positive sign, showing that banks extend more cross-border loans when bank funding conditions are accommodative. Although the effect of low interest rates on crossborder lending remain the subject of much debate, estimation results suggest that low rates increase the global banks’ risk-taking and accordingly the Turkish banks’ external debt. The US term premium has a negative coefficient, suggesting the presence of ‘search for yield’ incentives in global banks: Banks borrow short-term and lend long-term, so their domestic investment opportunities are less profitable when the yield curve is flat. This triggers banks’ search for yield and increases the Turkish banks’ cross-border loans. The M2 growth is also positively associated with the Turkish banks’ crossborder flows. Table IV.1.2 Estimation Results Dependent Variable: Independent Variables CBOE VIXt-1

The Logarithmic Value of Cross-Border Bank Loans (1) -0.00** (0.00)

TED Spreadt-1

(2)

(3)

(4)

(5)

(6)

-0.03* (0.02)

Slope of Yield Curvet-1

-0.00*** (0.00)

Real Policy Ratet-1

0.03*** (0.00)

M2 Growth Ratet-1

Deposit/Assets-1 Capital/Assets-1 Liquid Assets/Assets-1 NPLt-1 Profit/Assets-1 TR Macro Variables Real GDP Growtht-1 Inflationt-1 Real Effect. Exc Ratet-1 Constant Bank / Lender Country / Loan Types Fixed Effects Time Fixed Effects

(9) -0.00*** (0.00) -0.02 (0.02) 0.00 (0.00)

0.00*** (0.00)

Credit/GDP t-1

Credit/Assets-1

(8) -0.00** (0.00) -0.01 (0.02) 0.00 (0.00)

-0.11*** (0.03)

Real Credit Growtht-1

Bank-Specific Variables Log(Real Assets)t-1

(7)

0.01*** (0.00)

0.05** (0.02) 0.01 (0.01)

0.04* (0.02) 0.01 (0.01)

(10) -0.00* (0.00) -0.01 (0.02)

(11) -0.00*** (0.00) -0.01 (0.02)

-0.01 (0.04) 0.01*** (0.00)

-0.03 (0.04) 0.00 (0.00)

0.00 (0.01)

0.31*** (0.11) 0.01** (0.00) -0.03*** (0.01) -0.01 (0.01) -0.01*** (0.00) 0.00 (0.01) 0.02 (0.01)

0.40*** (0.01) 0.02*** (0.00) -0.02*** (0.00) 0.01*** (0.00) -0.00*** (0.00) -0.03*** (0.00) 0.01*** (0.00)

0.29*** (0.08) 0.01*** (0.00) -0.02*** (0.01) -0.00 (0.01) -0.01*** (0.00) 0.00 (0.01) 0.02** (0.01)

0.21*** (0.02) 0.01*** (0.00) -0.02*** (0.00) -0.00*** (0.00) -0.01*** (0.00) 0.00 (0.00) 0.03*** (0.00)

0.26*** (0.02) 0.01*** (0.00) -0.03*** (0.00) -0.00*** (0.00) -0.01*** (0.00) 0.00 (0.00) 0.02*** (0.00)

0.21*** (0.02) 0.01*** (0.00) -0.02*** (0.00) -0.01*** (0.00) -0.01*** (0.00) 0.00 (0.00) 0.03*** (0.00)

0.30*** (0.11) 0.01** (0.00) -0.03*** (0.01) -0.01 (0.01) -0.01*** (0.00) 0.00 (0.01) 0.02* (0.01)

0.31*** (0.11) 0.01** (0.00) -0.03*** (0.01) -0.01 (0.01) -0.01*** (0.00) 0.00 (0.01) 0.02 (0.01)

0.00 (0.00) 0.72 (0.44) -0.00 (0.00) 6.39*** (1.43)

0.00 (0.00) 0.61 (0.45) -0.00 (0.00) 6.37*** (1.43)

0.01*** (0.00) -2.96*** (0.47) 0.01*** (0.00) 4.16*** (0.10)

-0.01 (0.01) -0.99* (0.51) -0.00*** (0.00) 6.26*** (1.05)

0.01*** (0.00) -1.43*** (0.48) -0.00** (0.00) 7.33*** (0.34)

0.00** (0.00) 0.36 (0.51) -0.00** (0.00) 5.40*** (0.39)

0.00* (0.00) -1.80*** (0.47) -0.00 (0.00) 7.20*** (0.33)

8.86*** (0.36)

0.00 (0.00) 0.34 (0.43) -0.00 (0.00) 6.18*** (1.41)

9.47*** (0.31)

0.00 (0.00) 0.57 (0.44) -0.00 (0.00) 6.48*** (1.44)

Yes

Number of Observations 136,783 136,783 136,783 136,783 136,783 136,783 136,783 136,783 136,783 136,783 R2 0.14 0.14 0.07 0.14 0.14 0.14 0.14 0.13 0.14 0.13 ***, **, * indicate statistical significance at 1 percent, 5 percent and 10 percent, respectively. Values in parentheses refer to robust standard errors.

Yes 136,783 0.14

Since the correlation between individual US factors is moderate in our sample, we add most drivers simultaneously to the

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr model in regressions through 8 and 11.1 According to the 9th and 11th regressions, where the demand side as well as the fixed effects of the borrower bank, lender country, loan type and time are controlled by bank-based variables and macro indicators, the VIX representing the global risk appetite has significant and robust effects on Turkish banks’ cross-border loans at 1 percent level. In addition, we find a significant effect of global banks’ funding conditions on the external debt of Turkish banks.2 As seen in Table IV.1.2, the aforementioned global liquidity indicators have statistically significant effects on cross-border bank loans of the Turkish banking sector. This implies that sharp movements in global liquidity indicators as a result of financial shocks in any of the lender countries or banks have the potential to create volatility in cross-border loans of Turkish banks. At this point, we test whether the increased diversification in funding sources has had a limiting effect on the mentioned volatility. In Table IV.1.3, we include in the model not only global liquidity indicators but also their interactions with the number of lender countries that provide funds to borrower Turkish banks. When indicators are considered individually as in the first seven rows in the Table, the sensitivity of Turkish banks’ cross border loans to all global liquidity indicators, except the slope of the US yield curve, is lowered by the increasing diversification in lender countries/banks. According to the 8th and 9th regressions, where all indicators are added to the model simultaneously, the sensitivity of cross-border bank loans to the global liquidity indicators, namely the VIX index, TED spread, Federal Funds Target Rate and the US credit conditions, decreases as a result of the increasing diversification in lender countries/banks. Hence, the estimation results indicate that the Turkish banking sector has not only continued to roll over its debts from traditional financial centers but also gained access to new countries and banks. Moreover, the increasing diversification of lender origin limits the risks related to the concentration of lender countries and lowers the sensitivity of cross-border bank loans to global liquidity conditions. Due to the high correlation between the real policy rate representing the monetary policy and the slope of the yield curve, we do not add them to the model simultaneously. Similarly, due to the high correlation between real credit growth, which represents the lending appetites of global banks, and the ratio of private credits to GDP, we also do not add them to the model simultaneously. 1

We control the funding conditions of foreign global banks via TED spread, real credit growth and ratio of private credit to GDP. In the 9th and 11th regressions, while the effect of TED spread and real credit growth is insignificant, we find a significant and robust effect of the private credit to GDP ratio on crossborder bank loans. 2

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Table IV.1.3 Estimation Results Dependent Variable: Independent Variables CBOE VIXt-1 (Country*CBOE VIX)t-1

(1) -0.01*** (0.00) 0.00*** (0.00)

TED Spreadt-1

(2)

The Logarithmic Value of Cross-Border Bank Loans (3) (4) (5) (6) (7)

-0.19*** (0.04) 0.01*** (0.00)

(Country*TED Spread)t-1 Slope of Yield Curvet-1

0.00 (0.00) -0.00 (0.00)

(Country* Slope of Yield Curve)t-1 Real Policy Ratet-1

-0.09*** (0.02) 0.00* (0.00)

(Country*Real Policy Rate)t-1 Real Credit Growtht-1

Credit/GDP t-1

0.06* (0.03) -0.00** (0.00)

(Country*Credit/GDP) t-1 M2 Growth Ratet-1 (Country*M2 Growth Rate)t-1 0.01*** (0.00) 11.34*** (0.15)

Constant

0.01*** (0.00) 11.23*** (0.15)

0.01*** (0.00) 9.27*** (0.17)

0.01*** (0.00) 10.06*** (0.18)

0.01*** (0.00) 9.31*** (0.17)

(9) -0.01** (0.00) 0.00** (0.00) -0.08 (0.06) 0.00 (0.00)

-0.09* (0.05) 0.00*** (0.00) 0.01*** (0.00) -0.00** (0.00)

0.01*** (0.00) -0.00* (0.00)

(Country*Real Credit Growth)t-1

Countryt-1

(8) -0.01** (0.00) 0.00** (0.00) -0.15*** (0.06) 0.01*** (0.00) 0.00 (0.00) -0.00 (0.00)

0.01*** (0.00) 8.74*** (0.30)

0.04*** (0.01) -0.00*** (0.00) -0.00 (0.00) 6.57*** (0.34)

0.11*** (0.03) -0.00*** (0.00) 0.02 (0.01) -0.00 (0.00) 0.02*** (0.00) 8.35*** (0.32)

0.00 (0.01) -0.00 (0.00) -0.00 (0.00) 9.54*** (0.20)

Bank / Lender Country / Loan Types Yes Yes Yes Yes Yes Yes Yes Yes Yes Fixed Effects Time Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Bank-Specific Variables Yes Yes Yes Yes Yes Yes Yes Yes Yes TR Macro Variables Yes Yes Yes Yes Yes Yes Yes Yes Yes Number of Observations 136,727 136,727 136,727 136,727 136,727 136,727 136,727 136,727 136,727 R2 0.14 0.14 0.13 0.13 0.13 0.13 0.14 0.14 0.13 ***, **, * indicate statistical significance at 1 percent, 5 percent and 10 percent, respectively. Values in parentheses refer to robust standard errors.

IV.1.5

Conclusion

With the rise in international activities, financial integration has also deepened and gained strength globally, and capital flows from advanced countries to emerging countries have increased. The high sensitivity of cross-border bank loans to global liquidity developments have the potential to create volatility in capital flows as a result of financial shocks in any of the lender countries or banks. In this context, the transmission of financial shocks in a certain region to other countries through financial linkages, and the potential macrofinancial imbalances resulting from capital flow volatility have led borrowing countries to take measures against the negative effects induced by the volatility in capital flows. In an environment of macroprudential policies implemented by policy makers in Turkey, the sector continues to roll over its debts from traditional financial centers, but it has also gained access to new countries/banks. Accordingly, the number of lender sources that provide funds has increased steadily, and regions outside the traditional financial centers have also become important sources of funds for the Turkish banking sector. Estimation results indicate that the increasing diversification of lender countries/banks lowers the sensitivity of cross-

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr border bank loans to global liquidity conditions and therefore limits the spillovers of potential financial shocks that may occur in systemically important sources.

References

Cerutti, E., Claessens, S., & Ratnovski, L. (2014). Global liquidity and drivers of cross-border bank flows (No. 14-69). International Monetary Fund. Cetorelli, N., & Goldberg, L. S. (2011). Global banks and international shock transmission: Evidence from the crisis. IMF Economic Review, 59(1), 41-76. Lane, P. R. (2014, January). Capital flows in low-income countries. In Conference on â&#x20AC;&#x2DC;Macroeconomic Challenges Facing LowIncome Countries. IMF (2009, April). How linkages fuel the fire: the transmission of financial stress from advanced to emerging economies. World Economic Outlook, Chapter 4, 133-169.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Bankası Türkiyr

IV.2

Measures on Corporate Sector’s Access to Finance

Corporate sector finance is mainly provided through domestic bank loans. Since the second half of 2015 when the banking sector tended to tighten corporate loan standards, corporate loan growth has followed a relatively weakening trend. To break the cycle of weakening economic activity, increased credit risk and tightened credit standards that emerged in the face of a series of shocks experienced in 2016, prudential measures have been taken recently to support the access of the corporate sector, and SMEs in particular, to finance. Consequently, the increase in the guarantee limit of the Treasury-supported Credit Guarantee Fund (KGF), TOBB's low-interest loan facility (Respite Credit) and KOSGEB's interest-free loan support have contributed positively to corporate loan growth and also mitigated possible risks to economic growth that may come from the credit channel. Effective implementation of the aforementioned measures, which stand out with collateral and interest support, is important in terms of limiting downside risks to economic activity. In this regard, the mechanism of these measures and their implications for credit growth and interest rates are analyzed in this study.

IV.2.1

Recent Measures Taken to Increase Corporate Access to Finance Figure IV.2.1 Measures to Increase Corporate Access to Finance

Due to the close interaction between credit growth and economic activity, the effects of negative developments on the real economy may incrementally increase macroeconomic and fiscal vulnerabilities as a result of fiscal acceleration. In this context, prudential measures have been taken for the corporate sector, aiming to reduce risks to economic growth that may originate from the credit channel in the recent period.

Loan with KGF Guarantee Low loan interest rate Treasurybacked KGF guarantee Lower risk weight

Interest free

The foremost measure in terms of corporate financial access was the amendment focusing on the increase in the Treasurysupported KGF guarantee limit. The KGF supports financial access of companies with insufficient collateral by providing guarantee facility. Among the sources of KGF guarantee, Treasury support seems to be the main source besides the fund’s own equity and finance provided from abroad.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey

Treasury-Supported KGF Guarantee

The KGF guarantee limit supported by the Undersecretariat of Treasury has been increased to TL 250 billion from TL 20 billion.1 The Treasury compensation limit to be applicable in cases where the credit is not properly paid back in full and/or on time, i.e. if the credit with KGF guarantee is classified as a non-performing loan, was increased from TL 2 billion to TL 25 billion, which corresponds to 10 percent of the KGF guarantee.2 Following the publication of the Decree of the Council of Ministers regarding the increase of the Treasury-supported KGF guarantee limit to TL 250 billion, a protocol was signed between the Treasury and the KGF in March 2017. With this protocol, the guarantee volume was initially determined as TL 200 billion and the Treasury compensation limit for NPLs was set as 7 percent of guarantee.3

The maximum KGF guarantee/loan ratio for SME loans, corporate loans for exporters and companies performing foreign exchange (FX) earning activities (Eximbank and other banks) and other corporate loans has been increased to 90, 100 and 85 percent, respectively.4 Additionally, in order to reduce the cost of funding for KGF guarantee, guarantee commission fee was reduced to one time 0.03 percent from the 0.5 – 2 percent range, and the application fee was canceled. The Treasury-supported KGF guarantee can be used for TL, FX or FX-indexed credits. Working-capital loans can have a maturity of between 6 months and 5 years while investment loans can be used with a maturity of 6 months to 10 years5. On the other hand, loan interest rates are determined by banks.

While risk weights applicable in the capital adequacy calculation are 75 and 100 percent for SME loans and large scale corporate loans, respectively, the risk weight on TL loans used within the Treasury-supported KGF guarantee scheme can be taken as zero

Decision No. 2017/9969 Regarding the Amendment of the Treasury Support to the Credit Guarantee Institutions 1

Law on the Amendment of the Law on Retirement Fund of the Republic of Turkey, Certain Laws and Decree Laws 2

KGF news item headlined “Up to TL 200 Million KGF Guarantee to Companies”, http://www.kgf.com.tr/index.php/en/ ; newspaper item titled “KGF relieves bankers in terms of capital". 3

In the previous period, the guarantee / loan ratio was 85, 100, 85 and 75 percent in SME loans, export loans (Eximbank), export loans (other banks) and other corporate loans, respectively. 4

There is a maximum grace period of 1 year for working-capital loans and 3 years for investment loans.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Bankası Türkiyr percent under certain conditions. This practice positively affects the sector’s capital adequacy ratio for loans with KGF guarantee.

In addition to the KGF guarantee, banks may demand additional collateral from companies. The KGF, however, will be entitled to these collaterals to the extent of the guarantee/loan ratio. The additional collaterals demanded by the banks or the collections obtained through legal follow-ups are transferred to the KGF at the indemnified guarantee ratio.

Treasury support partially reduces banking sector credit risk. Following the determination of participant banks’ portfolio limit, the

Figure IV.2.2 Compensation Upper Limit for Treasury-Supported KGF Guaranteed Loans

KGF can pay, with Treasury support, up to 7 percent of the guarantee for the NPL portion of the loans provided within this limit.

For instance, assume that Bank A has a guarantee limit of TL 10 billion. In this case, Bank A may lend up to TL 10-12 billion of credit with

Guarantee / Loan Ratio (%)

SME 90

Export 100

Other 85

SME 90

Export Other 100 85 SME 90

SME 90

export etc. If one of the SME loans with Treasury-supported KGF

guarantee amount equal to 90 percent of the loan, whereas it may

SME 90

this guarantee depending on the segment of the credit such as SME,

guarantee turns into NPL, Bank A may receive from the Treasury a

Other 85

Treasury Compensation Cap

receive a guarantee amount as much as the loan amount when a credit extended as an export loan (both SME and large scale) turns into NPL. On a loan basis, for NPLs, banks can get compensation from the Treasury at the guarantee/loan ratio (85, 90 and 100 percent). However, this facility is valid until the NPL ratio reaches 7 percent for the loans the bank extends with the Treasury-supported KGF guarantee. If the NPL ratio exceeds 7 percent, the credit exposure above 7 percent will be beared by the bank. In short, the bank will be able to receive payment at the rate of guarantee if the NPL ratio for the bank-specific portfolio is less than 7 percent, whereas it must bear the increased credit risk if the NPL ratio exceeds 7 percent. In this context, it is important for the banking sector to perform effective credit risk management for the loan portfolio with Treasury-supported KGF guarantee.

The 7-percent upper limit of compensation urges banks to perform effective risk management. In this regard, the risk of loans with KGF guarantee to turn into non-performing loans is expected to evolve in a way similar to the existing loan portfolio.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Portfolio Guarantee System The "Portfolio Guarantee System" (PGS) has been developed within the KGF to speed up the credit utilization process. With the PGS, credit applications for up to TL 12 million in SME loans and up to TL 50 million in large scale company loans within the Treasurysupported KGF guarantee scheme are directly evaluated by the bank. Following the bank’s approval of the loan, the KGF only seeks compliance with the beneficiary conditions set out in the Decree of the Council of Ministers. This system contributes to getting quicker results for the loan application. For loans that exceed the abovementioned limits, in addition to the bank, the KGF also makes a credit evaluation. With the recent amendment, at least 80 percent of the total guarantee is aimed to be granted in the scope of PGS, especially as SME loans.

KOSGEB’s Interest-Free Loan Support

In December 2016, a credit facility with an upper limit of TL 50,000 and with a maturity of 36 months (and a grace period of the first 12 months), in which interest payments would be paid by KOSGEB, was introduced for KOSGEB member firms. The demand for this facility was fairly high, and 15,000 companies were able to benefit from these interest-free loans in the first phase. In the second phase in March, approximately 460,000 companies that fulfilled the application requirements and requested interest-free loan support were able to benefit from this opportunity. In the latter phase, the loan amount ranged between TL 20,000 and 50,000 depending on the firm size. This support aims to create a loan volume of approximately TL 11 billion mainly for micro and small businesses. In the case of KOSGEB interestfree loan support, the KGF guarantee can also be used if the company requests.

TOBB’s Low-Interest Respite Credit The Respite Credit has been designed in consideration of the KGF guarantee facility and TOBB’s deposits at Ziraat Bank and Denizbank. These two banks can grant loans to TOBB member companies with a maturity of 1 year and 9.9 percent annual interest rate. This scheme envisages creation of a loan volume of TL 5 billion and allocation of KGF guarantee of TL 4.2 billion for this project.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Bankası Türkiyr Currently,

wholesale

and

retail

trade,

manufacturing

and

construction sectors are at the forefront in the use of Respite Credit.

IV.2.2

Loan Growth and Interest Rate Developments Following the Measures

In the first four months of 2017, the acceleration in loan growth was driven by measures taken to increase corporate access to finance and support corporate loans. The strong loan growth

Chart IV.2.1 Corporate Loan Growth (Annual Percentage, FX Adj.)

particularly in the post-March 2017 period has mainly stemmed from

TOBB Respite Credit KOSGEB Interest Free Loan - I KOSGEB Interest Free Loan - II Increase in KGF Guarantee Limit

the increase in the Treasury-supported KGF guarantee limit along with companies’ and banks’ intense interest in this facility (Chart IV.2.1).

TRY

30 25

The KGF guarantee portfolio amount allocated to banks

reached about TL 160 billion as of 30 April 2017. Of this amount, TL 106-

billion credit volume (with a guarantee amount of TL 93 billion) was

granted to companies.6 Banks extend credits within approximately

one month following the credit approval by the KGF. Considering that the TL corporate loan stock increased by TL 105 billion from 30

Total

FX -5 11.16

12.16

01.17

02.17

03.17

04.17

Source: CBRT (Latest Data: 28.04.17)

December 2016 to 28 April 2017, it can be said that the KGF guarantee facility has been used extensively. The KGF guarantee is mainly used in TL denominated and SME loans. In terms of loan type, working capital loans that can be used with a maturity from 6 months to 5 years stand out. Chart IV.2.2 Corporate Loan Growth By Scale

growth

accelerated

February

and

March

(Chart

IV.2.2).

Considering that loans with the KGF guarantee have been used predominantly since mid-March, the strengthening in SME loan growth is expected to continue for a while.

02.17

16 14 12 10 8

+1.8 15.2 17.0 17.5 18.8

determining role in TL corporate loan growth as of 2016, SME loan

12.16 01.17

18 +2.4

03.17

+3.3 8.4

period. As a matter of fact, while large scale corporate loans had a

6 4.6 5.1 6.1

denominated SME loan growth may accelerate in the upcoming

(Annual Percentage, FX Adj.)

10.9 12.3 12.9 14.6

Due to the fact that measures focus mainly on SMEs, TL-

4 2 0 Total Corporate

SME

Large Scale

Source: CBRT (Latest Data: 03.17)

Despite the increase in the SME loan usage, interest cost for these loans remained at reasonable levels. Interest rates for large and medium-scale corporate TL loans gradually increased in 2017, while there is a more favorable structure in micro and small-scale corporate 6

Including KOSGEB interest-free loan support (upon applicant-company’s KGF guarantee request) and TOBB Respite Credits (direct).

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey

Chart IV.2.3

loan rates in which credit support is intensified (Chart IV.2.3). It is

Corporate Loan Interest Rate by Scale1 (Excluding credit cards and overdraft accounts, 4-week MA, Percentage)

estimated that TOBB and KOSGEB credit supports were the

SME Micro Small Medium Large Scale

determinant of the decline in micro-scale corporate loan rates in January and March.

While loan growth has recovered thanks to the regulatory 14

amendments effective since 2016Q3, the KGF guarantee facility has also supported this recovery particularly from mid-March onwards. Treasury-supported

03.17

12.16

09.16

06.16

03.16

12.15

KGF

guaranteed

loans

are

predominantly

provided in the form of new loans, which has been influential in this

1) Excluding zero percent interest rate loans for large scale companies. Source: CBRT (Latest Data: 28.04.17)

development.

Chart IV.2.4

Along with the recent acceleration in loan growth, mainly FX

Loan Usage and Funding Structure between December 2016 – April 2017 (Billion TRY) 1

deposits increased on the funding side. Compared to 2016 year-end, the increase in TL corporate loans amounted to TL 105 billion while TL

100

deposits grew by TL 4 billion and FX deposits rose by TL 83 billion. The

Increase in TL Stock Corporate Loans

Increase in TL Deposits

Increase in FX Deposits

Info: Net TL Swap Funding; 116

rise in FX deposits was reflected in TL currency swaps (Chart IV.2.4).

Increase in External Debt, 9

Increase in FX Deposits; 83

Increase in TL Stock Corporate Loans; 105

Increase in TL Deposits; 4

120

Increase in External Debt

Loan interest rates for SMEs followed a flat course until the end of April and the TL commercial loan-deposit rate spread narrowed slightly, while commercial deposit and currency swap rates increased

Info: Net TL Swap Funding

(Chart IV.2.5).

1) December 30, 2016 – April, 28 2017 period develomnets.

Source: CBRT (Latest Data: 28.04.17)

IV.2.3

Conclusion

Chart IV.2.5 Deposit and 3 Month Currency Swap Interest Rates (4-week MA, Percentage)

TL Corp. Loan - Deposit Interest Spread TL Corporate Deposit Interest Currency Swap Rate

15 13

Treasury-supported KGF guarantee, the TOBB Respite Credit and the

KOSGEB interest-free credit facility. Both banking and corporate

sectors have intense interest in these mechanisms, and there has

03.17

12.16

09.16

06.16

03.16

12.15

09.15

06.15

03.15

12.14

Source: CBRT (Latest Data: 28.04.17)

been a strong increase in corporate loan growth in the recent period.

Following the protocol between the Treasury and the KGF, loans

granted via the KGF guarantee facility accelerated also due to

companies’ retained loan demand. These credits’ low risk weight in

7 5

Corporate sector access to finance has been supported by the

capital adequacy calculations and the strong collateral structure have bolstered utilization of credit with KGF guarantee. Credit utilization is expected to follow a more balanced path throughout the year after KGF-guaranteed credit demand reaches a certain degree of saturation.

Financial Stability Report – May 2017

Centra Central Bank of the Republic of Turkey Türkiyr

Türkiyr

IV.3

Drivers of Credit Dollarization

Abstract

While the tendency of banks to match their foreign currency (FC) assets with their FC liabilities feeds the supply side of credit dollarization, firms’ stronger preference for FC credits, relying on their natural heghes (e.g. export revenues), supports the the demand side of credit dollarization. Findings of this special topic show that both (bank and firm) tendencies increase credit dollarization while the bank side effect is stronger. Among bank liabilities, non-core FC liabilities (such as FC bonds, syndications and securitization) were found to be a stronger driver of credit dollarization compared to FC deposits.

IV.3.1

Introduction

Since the global financial crisis, increased capital flows supported by the expansionary monetary policies of central banks of developed countries have steadily fed the credit dolarization in developing countries.1 These FC credits are attractive for firms as they allow access to cheaper finance with longer maturities. However, they contain the risk of deteriorating firms’ financial structures, particularly at times of high exchange rate volatility.2 As a matter of fact, these negative effects can be critical for the stability of the financial system. In this special topic, the effect of bank and firm tendencies, the two sides of the loan supply and demand relationship, on feeding credit dollarizaiton is examined. In general, concentration of deposits in FC as a result of volatility in local currency; international funds being available mostly in hard currencies (e.g., USA dollars, euros)

generate

significant

dollarization

bank

liabilities in

developing countries (Hausmann et al. 2001)). Lending these FC

Hake et al. (2014) provides a detailed meta-analysis of the credit dollarization and its causes. Honohan (2006) discusses dollarization trends and interaction with macro variables for Turkey. The distribution of the recent firm level credit dollarization trends in Turkey is studied descriptively in Hülagü and Yalçın (2014). 2

Barajas et al. (2016) analyze the data from Colombian firms that shows that the financial structures of firms with high FX credits have deteriorated during the sudden devaluation of the local currency and their investment performance has significantly weakened. They also emphasize that only a small part of this deterioration can be recovered by the exchange rate recovery.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Türkiyr funds directly in local currency is restricted by regulations, as that would expose banks to currency risks.3 For this reason, banks in Turkey can only extend loans from these FC funds in TL, once they have bought protection against exchange rate risk through swap transactions. Alternatively, the banks can transfer the currency risk to borrowing firms by directly lending in FC (e.g. asset-liability matching).4 On the demand side, firms with natural hedges (e.g. export revenues) against exchange rate fluctuations prefer low-cost FC loans, particularly in long-term financing.5 Firms’ tendency towards low-cost FC loans, relying on their natural hedges and banks’ asset-liability matching propensity constitute the demand and supply sides of credit dollarization. The results of this analysis show that both firms’ and banks’ behaviors feed credit dollarization. When the size of the effects is examined, the effect of banks' asset-liability matching tendency appears to be stronger. It is interesting and important that the effects of the above mentioned tendencies on credit dollarization are not linear. In other words, while the increase in natural hedges (FC incomes) raises the credit dollarization regressively, the increase in banks' FC liabilities supports the credit dollarization progressively. A look into the structure of banks' FC liabilities indicates that the effect of FC deposit and non-core FC liabilities on credit dollarization is different. According to our findings, while non-core FC liabilities (e.g. FC

bonds,

syndications

and

securitizations)

are

transformed

significantly into FC loans, the effect of FC deposits with relatively low maturities on credit dollarization is considerably limited. The results of the study are generally in line with the related literature. For example, Brown et al. (2014) concluded that the bank that was the subject of their analysis, performed asset-liability matching by increasing the FC credit acceptance rates during periods of increased FC liabilities. Luca and Petrova (2008) also support this finding with their analysis of aggregate panel data. Alp and Yalçın (2015) and Özsöz and others (2015) are the most striking

According to the current regulations, the net FC position (FC asset - FC liability) of banks in Turkey can not exceed twenty percent of risk weighted assets (Official Gazette No: 26333dated November 2006 ). Currently, this ratio is much lower than the regulated limit for many banks. 4

As per current regulations, real persons in Turkey can not borrow in FC. Firms with FC revenues or large firms with no FC revenues but have a capacity to borrow over 5 million USD dollars can obtain FC loans (domestic FC loans). Firms that do not have any FC income are only permitted to obtain FC indexed loans (CBRT 2009 / YB-22). 5

The FC financing cost here does not include the hedging cost. As a matter of fact, firms' FC revenues already provide natural protection.

Financial Stability Report – May 2017

Centra Central Bank of the Republic of Turkey TĂźrkiyr ones among the papers using Turkish data. The first one, which only TĂźrkiyr controls for the firm-side effects, primarily relates the credit dollarization with firmsâ&#x20AC;&#x2122; FC revenues, while the second one only focuses on the supply side (banks) and, concludes that credit dollarization is largely backed by banksâ&#x20AC;&#x2122; FC liabilities. In our current work, both sides of the loan relationship are controlled for by taking into account the demand and supply sides at the same time. In this regard, it is the first study on Turkey which considers the both sides.

IV.3.2

Estimation Methodology

The model presumes that both banks and firms are risk averse and they adopt the minimum variance portfolio (MVP) method in their lending and borrowing decisions. While the basic dynamics of this method were discussed in Ize and Levy-Yeyati (2003), Luca and Petrova (2007) established an empirical application of it. In this study, following the analysis of Luca and Petrova (2007) that uses country level aggregate data, the following model is estimated by incorporating firm and bank heterrgeneity:

đ?&#x2018;&#x152;đ?&#x2018;&#x192; đ??śđ?&#x2018;&#x;đ?&#x2018;&#x2019;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018; ( ) = â&#x2C6;?â&#x20AC;˛ (đ??šđ?&#x2018;&#x2013;đ?&#x2018;&#x;đ?&#x2018;&#x161;)đ?&#x2018;&#x2013;đ?&#x2018;Ąâ&#x2C6;&#x2019;1 + đ?&#x203A;˝ â&#x20AC;˛ (đ?&#x2018;&#x2020;đ?&#x2018;Śđ?&#x2018;&#x203A;đ?&#x2018;Ąâ&#x201E;&#x17D;đ?&#x2018;&#x2019;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;? đ??ľđ?&#x2018;&#x17D;đ?&#x2018;&#x203A;đ?&#x2018;&#x2DC;)đ?&#x2018;&#x2013;đ?&#x2018;Ąâ&#x2C6;&#x2019;1 + đ?&#x2018;&#x2021;đ?&#x2018;&#x153;đ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ??śđ?&#x2018;&#x;đ?&#x2018;&#x2019;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018; đ?&#x2018;&#x2013;đ?&#x2018;Ą đ?&#x153;&#x192; â&#x20AC;˛ (đ??śđ?&#x2018;&#x;đ?&#x2018;&#x2019;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ą đ?&#x2018;&#x192;đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;Ąđ?&#x2018;&#x201C;đ?&#x2018;&#x153;đ?&#x2018;&#x2122;đ?&#x2018;&#x2013;đ?&#x2018;&#x153;)đ?&#x2018;&#x2013;đ?&#x2018;Ąâ&#x2C6;&#x2019;1 + đ?&#x2018;&#x201C;đ?&#x2018;&#x2013; + đ?&#x2018;&#x2018;đ?&#x2018;Ą + đ?&#x2018;&#x2019;đ?&#x2018;&#x2013;đ?&#x2018;Ą The dependent variable is the ratio of firm iâ&#x20AC;&#x2122;s FC credits (to the synthetic bank that is defined specifically for firm i) to total credits (FC + TL) in date (yyyymm) t.6 The independent variables are the firm variables that are obtained from the annual balance sheet and income tables; the synthetic bank monthly balance sheet and income table variables, obtained by weighting bank variables using each bankâ&#x20AC;&#x2122;s share in the credit portfolio of firm i and finally, the characteristics of the firm-synthetic bank credit portfolio. In more detail, the synthetic bank can be expressed as the sum of the multiplication of each bank's share (đ?&#x203A;žđ?&#x2018;&#x2013;đ?&#x2018;&#x2014;đ?&#x2018;Ą ) in firm iâ&#x20AC;&#x2122;s credit portfolio at date t with its own balance sheet and income table. In this way, all bank relationships of firm i at time t are reduced to one synthetic bank, which allows our analysis to concentrate directly on the firmlevel credit dollarization. đ??˝

(đ?&#x2018;&#x2020;đ?&#x2018;Śđ?&#x2018;&#x203A;đ?&#x2018;Ąâ&#x201E;&#x17D;đ?&#x2018;&#x2019;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;? đ??ľđ?&#x2018;&#x17D;đ?&#x2018;&#x203A;đ?&#x2018;&#x2DC;)đ?&#x2018;&#x2013;đ?&#x2018;Ą = â&#x2C6;&#x2018; đ?&#x203A;žđ?&#x2018;&#x2013;đ?&#x2018;&#x2014;đ?&#x2018;Ą [đ??ľđ?&#x2018;&#x17D;đ?&#x2018;&#x203A;đ?&#x2018;&#x2DC; đ??ľđ?&#x2018;&#x17D;đ?&#x2018;&#x2122;đ?&#x2018;&#x17D;đ?&#x2018;&#x203A;đ?&#x2018;?đ?&#x2018;&#x2019; đ?&#x2018;&#x2020;â&#x201E;&#x17D;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;Ą]đ?&#x2018;&#x2013;đ?&#x2018;&#x2014;đ?&#x2018;Ą đ?&#x2018;&#x2014;

Since, detailed data on the FC credits obtained from foreign financial institutions abroad (cross boarder lending) is not available, the analysis only includes domestic FC and FC indexed credits.

Financial Stability Report â&#x20AC;&#x201C; May 2017

Central Bank of the Republic of Turkey Türkiyr All the independent variables used in the model are presented in Table IV.3.1. Firms’ natural hedge is captured by "export / total sales" and banks’ liability dollarization is measured by "FC liability / total liabilities" variables. Besides, liability dollarization was examined in two different parts: deposits (FC deposit / total liabilities) and non-deposit FC liability (non-deposit FC liability / total liability) dollarization. Table IV.3.1 Independent Variables Firm Variables

Synthetic Bank Variables

Credit Portfolio Variables

Log (Tot. Assets)

(w) Log (Tot. Assets)

Main Financing Sector

Export/ Tot. Sales

(w) FC Liabilities/Tot. Liab.

Num. Of Credit Sector

Tengiable Assets./ Tot. Assets

(w) FC Deposits/Tot. Liab.

Share of Midterm Maturity (12-24 Months) (5)

Trade Credits/Tot. Debt (1)

(w) Deposits Excl. FC Liab /Tot. Liab.

Share of Longterm Maturity (24+ Months)

Tot. Debt/ Tot. Assets (1)

(w) Tot. Liab./ Tot. Assets

Interest Coverage Ratio (ICR) (2)

(w) Profits/Equity (w) NPL/(NPL+Tot. Credits) (3) Synthetic Bank Ownership (4) Main Financing Bank Num. Of Financing Bank (in Sythetic Bank)

(1)Total Debt: Sum of Financial, trade ve other credits. (2) ICR EBITA/ Finance Costsi. (3) All (houcehold and firm) credits. (4) Ownership Public, private, foreign, participation etc. (5) Contains all (TL and FC) credits blonging to firm i. (w) Weighted (synthetic bank) variables. Deposits Excl. FC Liab: FC assets - liabilities.

Moreover, the factors that may affect firms' FC loan demand are selected as independent variables by following the related literature. In short, these include: firm size (total assets), firm capital (investment) status (share of tangible assets in total assets), access to non-bank financial facilities (trade credits as a share of total liabilities), general leverage (total debts / total assets) and financial strength (interest coverage ratio). Likewise, the size of the (synthetic) bank, the leverage, equity profitability, the risk appetite in the credit portfolio (NPL ratio), ownership ratios (e.g. public and private), the number of banks entering the synthetic bank total and share of the bank with the highest share in the synthetic bank – main financing bank (MFB). Finally, the share of the "main credit sector (subject)" (MCS) with the largest share in terms of the subject of the loan portfolio, the total number of credit sectors obtained by firm i on each date and the maturity structure of the loan portfolio is controlled for the medium and long term. In order to deal with the potential endogeneity problems in the model, yearly firm variables are lagged by one year, monthly (synthetic) bank and loan portfolio variables are also lagged for one month. Besides controlling for the firm, MFB, MCS and time fixed effects, MCS is also interacted with time fixed effects in all

Financial Stability Report – May 2017

Centra Central Bank of the Republic of Turkey Türkiyr specifications. In addition to these, number of different specifications Türkiyr are performed to test the robustness of the main results, including interacting firm and MFB fixed effects with the time dummies. Interacting firm FE with time dummies allows us to fully control for the firm (demand) side and interacting MFB FE with time dummies also allows us to control for most of the supply side effects. One potential weakness of the estimation here is that a dynamic data generating process is not considered in the model despite the fact that the dependent variable is characterized based on stock (credit balance) data instead of a flow (loan) data. In order to account for this, the model is re-estimated with quarterly and annual dependent variables. The results presented below have remained substantially robust across all these different specifications.

IV.3.3

Data and Descriptive Statistics

Graph IV.3.1 Average Firms’ Export Share (Annual) and Average FC Credit Share (Monthly) (percent) 45

FC Credit Share

Exports/Tot. Sales (Right A.)

29 28

obtained from the CBRT's firm data base, monthly bank financial

data come from the BRSA and finally, monthly firm-bank matched

credit balance data was obtained from the TBB Risk Center (TBB RM). Descriptive statistics of only the variables that are of particular

11.06 05.07 11.07 05.08 11.08 05.09 11.09 05.10 11.10 05.11 11.11 05.12 11.12 05.13 11.13 05.14 11.14 05.15 11.15

The annual company balance sheet and income tables were

Source: CBRT (Latest Observation: 12.15)

importance for the purpose of this special topic are presented in Table IV.3.2. In general, firms’ tendency towards credit dollarization has declined over time and reached 31 percent by the end of 2015 (Chart IV.3.1). A similar trend is observed in firms’ average export shares in the same chart. Although the size of the relationship has been somewhat reduced since the FC lending regulation in 2009 7, the fact that they follow similar trends over time indicates that there is a strong relationship between credit dollarization and export sales (over 80 percent correlation).

Graph IV.3.2 Amount Distribution FC Credit Balance, as of December 2015 (percent) 1 M TL üzeri 100 M TL - 1 M TL

Considering the fact that the average firm size in the sample is

50 M TL - 100 M TL

96.8 million TL, there appears to be a bias towards large firms. Even

10 M TL - 50 M TL

though this may suggest that small firms are under-represented in the

5 M TL - 10 M TL

sample, it will have a limited impact on the results. This is mainly

1 M TL - 5 M TL

because FC credits are highly concentrated in large firms due to

100 B TL - 1 M TL

current regulations. In Graph IV.3.2, we compare credit balance distributions of FC loans with the corresponding population of TBB RM

Population Sample

0 TL - 100 B TL 0

Source: TBB Risk Center, CBRT

With the Council of Ministers decision No: 27260 dated June 2009, large firms with no FC revenues but have the capacity to borrow over 5 million USD dollars were permitted to borrow in FC with at least one year or longer maturity. In accordance with this decision, FC loans issued under this regulation are presumably used more actively in the financing of infrastructure projects. This may weaken correlation between export revenues and FX borrowing.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Türkiyr data for December 2015. According to the graph, the difference Graph IV.3.3

between the sample and the population is particularly minimal

Average Bank FC Liabilities Share and Average FC Credit Share (Yüzde)

especially at and above 5 million TL.8

55 FC Credit Share

44 42

FC Liab. /Tot. Liab (Right A.)

Similarly, the relation between the share of banks’ FC liabilities

in total liabilities and the firm level average credit dollarization is 45

presented in Chart IV.3.3. Although the relation appears to diverge

38 36

from the beginning of 2013 and on, generally a positive relationship

is observed. The correlation before 2013 was around 60 percent,

34 35

which

becomes

negative

2013.

Considering

different

components of banks' FC liabilities (FC deposits, syndication, etc.),

11.06 05.07 11.07 05.08 11.08 05.09 11.09 05.10 11.10 05.11 11.11 05.12 11.12 05.13 11.13 05.14 11.14 05.15 11.15

there appears to be a stronger correlation between the non-deposit

Source: CBRT (Latest Observation: 12.15)

FC funds and the credit dollarization (Chart IV.3.4). As a matter of fact, while FC deposits have been more stable, recently the share of non-deposit FC liabilities has significantly increased. Table IV.3.2 Summary Statistics Variables Dependent Variable (Monthly) FC Credits/Total Credits Firm Independent Variables (Annualy) Total Assets Export/Total Sales Synthetic Bank Independent Variables (Monthly) (w) Total Assets (w) FC Liabilities/Tot. Liab. (w) FC Deposits/Tot. Liab. (w) Deposits Excl. FC Liab /Tot. Liab.

IV.3.4

Observation

Mean

Std. Deviation

1,105,951

0.34

0.42

101,137 99,885

96,800,000 0.15

575,000,000 0.27

1,105,951 1,105,951 1,105,951 1,105,951

85,200,000 0.43 0.25 0.18

56,300,000 0.08 0.05 0.09

Empirical Findings

Graph IV.3.4 Shares of FC Liabilities in Banks’ Total Liabilities and Average Firm FC Credits Share (percent) 46

FC Credit Share FC Dep./Tot. Liab. (Right A.) Dep. Excl. FC Liab./ Tot. Liab (Right A) FC Liab./ Tot. Liab.

Summary of important empirical findings is presented in Table 55

IV.3.3. Findings show that firms’ propensity to rely on their natural

protections and the banks' asset-liability matching tendencies are

40 40

obvious drivers of credit dollarization (Table IV.3.3, columns (1) and

(2)).

According to the estimates, one standard deviation (0.27)

increase in firms' export/total sales ratio results in an average

38 36

20 15

11.06 04.07 09.07 02.08 07.08 12.08 05.09 10.09 03.10 08.10 01.11 06.11 11.11 04.12 09.12 02.13 07.13 12.13 05.14 10.14 03.15 08.15

Source: CBRT (Latest Observation: 12.15)

increase of 1.3 percentage points in firm level credit dollarization on a monthly basis, while the same increase in the banking sector (FC liability share) results in a 2 percentage point increase in credit dollarization. Comparing the standardized coefficient estimates shows that banks' tendencies feed credit dollarization more strongly than those of firms.

For more information on the concentration of FC credits among large firms, see: Financial Stability Report May and November 2016. 8

Financial Stability Report – May 2017

Centra Central Bank of the Republic of Turkey Türkiyr The effect of one standard deviation increase in banks' FC Türkiyr core liabilities, share of FX deposits in total liabilities, on credit dollarization is 1 percentage point, while for non-deposit FC liabilities, this effect was estimated at 2.2 percentage points. These findings show that the non-core FC funds obtained through bond issuance, syndication or securitization from abroad feed domestic credit dollarization more than core FC funds. This can be explained by the fact that non-core FC funds have a longer maturity structure than core FC funds and thus, are preferred more strongly in FC loan financing.

Table IV.3.3 Coefficient Estimates, Dependent Variable: FC Credits/Total Credits (1)

Export/ Tot. Sales

(2)

(4)

(5)

(6) At Least Once & Manufacturing 0.0741*** (0.0148)

All

At Least Once & Manufacturing

0.0497*** (0.0109)

0.127*** (0.0281)

0.0742*** (0.0148)

-0.0919*** (0.0314)

-0.0925*** (0.0314)

(Export/ Tot. Sales)2

(w) FC Liabilities/Tot. Liab.

(3)

0.249*** (0.0249)

0.0505 (0.0784)

(w) (FC Liabilities/Tot. Liab.)2

0.367*** (0.0428)

0.206** (0.0811)

(w) FC Deposits/Tot. Liab.

0.224*** (0.0385)

0.100 (0.162)

(w) (FC Deposits/Tot. Liab.)2

0.304*** (0.0722)

0.263 (0.303)

(w) Deposits Excl. FC Liab /Tot. Liab.

0.253*** (0.0263)

0.152*** (0.0444)

(w) (Deposits Excl. FC Liab /Tot. Liab.)2

0.371*** (0.0434)

0.130** (0.0616)

Other Firm Variables

Yes

Other (w) Bank Variables

Yes

Credit Portfolio Variables

Yes

791,866

300,277

0.841

0.781

Num of Observation R2

Robust (clustered) standard errors in parantheses, *** p<0.01, ** p<0.05, * p<0.1, (1) Firm variables are lagged for 12 months, and bank variables are also lagged for 1 month. (2) variable x and its square (x2) are jointly statistically significant according to all conventional levels. At Least Once: Firms that have used FC loans at least once during the sample period. Manufacturing: Main sector of credit is manufacturing sector.(w):Weighted synthetic bank variable. Deposits excluded YP Liabilities: FC Liabilities-FC Deposits. The firm, the main financier bank, the main credit sector, the date, and the main loan, are each checked on the basis of fixed variables.

In Table IV.3.3 columns (3) and (4), we add non-linear variables to our estimates and examine the increase or decrease rates of the above-mentioned effects relative to their initial levels. According to the results, the effect of firms’ natural hedges on the FC borrowing is positive but at a diminishing rate, whereas the reverse is observed on the bank side. In short, although firms prefer to borrow in FC by relying on their natural protection (e.g. export revenues), this trend is weakened as export levels increase. Such differentiation in the general tendency is considered to be the result of firms needing TL financing, in order to maintain their domestic

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey TĂźrkiyr production activities and to obtain short term financing. On the banks side, the effect of the asset-liability matching tendency on credit dollarization is generally increasing. In other words, the increase in the share of FC in total liabilities supports the credit dollarization at an increasing rate. The increasing effect of a rise in non-core FC funds (among FC liabilities) on the credit dollarization is faster than the increase in FC deposits. In sum, the (asset-liability matching) tendency of banks to issue FC loans from additional nondeposit FC liabilities is stronger when the level of non-deposit FC liabilities is high. Another important result from the analysis is that firm and bank trends are stronger in the manufacturing industry, which holds almost all of the country's exports (Table IV.3.3, columns (5) and (6)). As a matter of fact, coefficient estimates for the variables that measure bank and firm tendencies increase significantly, when only the firms with the main credit topic of manufacturing industry are considered. The effect of one standard deviation increase in firmsâ&#x20AC;&#x2122; natural hedges on credit dollarization reaches 2 percentage points, while a similar increase in bank FC liabilities leads to a rise of 3 percentage points. In support of the above findings, the effect of one standard deviation increase in non-deposit FX liabilities on credit dollarization is 3.2 percentage points, while the effect of a similar increase in FX deposits is 1.4 percentage points. The robustness of the findings is tested with different specifications against endogeneity due to potential unobserved factors affecting bank and firm behavior. Macro trends that affect both firms and banks similarly (e.g., uncertainty, monetary policy and tightening/expansion of financial conditions) are already controlled in all specifications with year-month dummies. Moreover, using the interaction of firm FE with time dummies, all observable or unobservable factors influencing firm behavior are examined in different specifications. Even in such a detailed specification, no significant variation was recorded in the coefficient estimates of bank variables. With a similar approach, the bank side was controlled through the interaction of time dummies with the main financing bank in each synthetic bank. In this specification, the firmside coefficient estimates remained largely the same. Finally, estimates have been repeated quarterly and annually to eliminate dynamic effects (instead of monthly bank variables). The results were qualitatively similar.

Financial Stability Report â&#x20AC;&#x201C; May 2017

Centra Central Bank of the Republic of Turkey Türkiyr

IV.3.5

Results and Policy Discussion

Türkiyr In this study, the effect of firm and bank behavior on credit dollarization is examined. Banks tend to reduce their exposures to exchange rate risk by matching FC liabilities with assets, and firms tend to prefer low cost and long term FC loans relying on their natural hedges. The analysis shows that these tendencies of banks and firms significantly feed credit dollarization. A comparison of the size of the impacts of both tendencies reveals that banks' assetliability matching behavior feeds credit-dollarization more strongly. The core FC liabilities of banks (FC deposits) seem to have a relatively limited impact on credit dollarization compared to other foreign sources (such as FC bond issuance, syndications and securitization).

References

Alp, Bengü and Yalçın, C (2015). Liability Dollarization and Growth Performance of Non-Financial Firms in Turkey. Working Paper, CBRT 15/01. Barajas, Adolfo; Restrepo, Sergio; Steiner, Roberto; Medellín, Juan Camilo and Pabón, César (2016). Balance Sheet Effects in Colombian Non-Financial Firms. Hake, Mariya and Lopez-Vicente, Fernando & Molina, Luis (2014). Do the Drivers of Loan Dollarization Differ between CESEE and Latin

America?

A Meta-Analysis. Focus on European

Economic Integration, Oesterreichische Nationalbank . Honohan, P (2006). Dollarization: Consequences and Policy Options, Prepared for the Central Bank of Turkey 75th Anniversary C. Hülagü, T and Yalçın, C (2014 Micro Level Evidence on Foreign Exchange Liability and the Exchange Rate Risk in Turkish Corporate Sector, CBRT Research Notes in Economics 13/25. Luca, Alina and Petrova, Iva, (2008). What drives credit dollarization in transition economies? Journal of Banking & Finance, Elsevier, 32(5), 858-869. Ozsoz, Emre; Rengifo, Erick W. and Kutan, Ali, (2015). Foreign Currency Lending and Banking System Stability: New Evidence from Turkey, Central Bank Review, 15, Sayı 2. Brown, Martin; Karolin Kirschenmann and Steven Ongena, (2014). Bank Funding, Securitization, and Loan Terms: Evidence from

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Türkiyr Foreign Currency Lending. Journal of Money, Credit and Banking, Blackwell Publishing, Sayı 46(7). Ize, Alain and Levy Yeyati, Eduardo, (2003). Financial dollarization, Journal of International Economics, 59, Sayı 2, 323-347. Hausmann, R.; U. Panizza, and E. Stein (2001). Why Do Countries Float the Way They Float? Journal of Development Economics, 66:2, 387-414.

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Türkiyr

IV.4 Effects of Retail Loan Regulations While prudent borrowing tendency of households has been maintained, macroprudential

standards

have

been

loosened

Chart IV.4.1 Retail Loans/GDP and Deviation from Trend (Percent)

Deviation from Trend (RHE) Retail Loans / GDP Trend

recently to some extent (Chart IV.4.1 and IV.4.2). Effective from 27 September 2016, the general installment limit for individual credit

cards was raised to 12 months, whereas the general purpose loan

cap was increased to 48 months. In addition, a restructuring facility,

which allows for a maturity of up to 72 months for the restructuring of

general purpose loans and individual credit card debt, was

-4

introduced. The loan to value ratio for housing loans was increased

-6

lowered. This study elaborates on macroprudential policy changes

4 2

12.16

12.15

12.14

12.13

12.12

12.11

12.10

12.09

12.08

12.07

12.06

12.05

12.04

12.03

-2

12.02

moderately. Meanwhile, general provision ratios for retail loans were

Note: HP filter method is used (λ=400.000).

Source: CBRT (Latest Data: 03.17)

since 27 September 2016 and their effects on retail loans. Chart IV.4.2 Household Financial Assets and Liabilities

IV.4.1

Regulations Regarding General Purpose Loan Maturities and the Individual Credit Card (ICC) Installment Limit

(Billion TL)

1.200

Assets Net Assets

1.000

Liabilities

800 600 400

01.17

07.16

01.16

07.15

01.15

maturity of about 75 percent of the new loans extended following

-600

07.14

general purpose loans has lengthened thereafter (Chart IV.4.3). The

-400

01.14

September 2016 (Table IV.4.1). The weighted average maturity of

0 -200

07.13

36 months since 2013 year-end, was raised to 48 months on 27

200

01.13

The maturity cap on general purpose loans, which had been

Source: CBRT, BRSA, CMB, MKK, TOKİ (Latest Data: 03.17)

this regulation is longer than 36 months. The facility of restructuring of general purpose loans with up to 72 month-maturity is also believed Chart IV.4.3

to have been influential in the maturity extension.

Average Maturity of General Purpose Loans and ICC (Stock, Months)

Table IV.4.1 Consumer Loan Maturity Limit

(Months)

31.12.2013

25.11.2015

27.09.2016

General Purpose Loans (in general)

Limitless

-Used for financing of education

Limitless

-In the context of house modification, purchases of goods and services as an integral part of the house

Limitless

Loan Type

01.12 05.12 09.12 01.13 05.13 09.13 01.14 05.14 09.14 01.15 05.15 09.15 01.16 05.16 09.16 01.17

Before

Effective Date

General Purpose ICC (RHE) 27.09.2016

Source: CBRT (Latest Data: 03.17)

Vehicle Loans

Limitless

Housing Loans

Limitless

Source: CBRT

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Table IV.4.2 ICC Installment Limits (Months)

Before

01.02.2014

13.05.2014

22.10.2014

25.11.2015

27.09.2016

Effective Date

Goods and services purchases and cash advances

Limitless

Intangible goods purchases

Limitless

Telecommunication, meal, food and fuel

Limitless

Jewelry purchases

Limitless

White goods, furniture, education

Limitless

Electronic goods and computer purchases

Limitless

Airlines, travel agencies, transportation, accommodation, health and social services, health product purchases, club and association payments, tax payments

Limitless

Purchases related to direct marketing, overseas purchases and alcoholic drinks, cosmetics and office equipment expenditures

Limitless

Sector

Source: CBRT

While the individual credit card (ICC) general installment limit was increased from 9 to 12 months; the limit was kept at 12 months for white goods, furniture and education expenditures and 4 months for jewelry purchases. The installment limit for electronic appliances and computer purchases was lowered from 9 to 6 months, whereas the existing 9-month limit for airlines, travel agencies, transportation, accommodation, health expenditures and tax payments was preserved. While the installment ban on telecommunication, meal, food, alcohol and fuel oil expenditures remained intact, overseas and cosmetics expenditures were also classified as non-installment expenditures in the new regulation (Table IV.4.2). These changes to sectoral installment limits that contribute to the current account balance and seek a match Chart IV.4.4

between the periods of consumption and financing also aim to

Change in Credit Card Installment Cap and Installment Expenditures by Sectors

produce balanced consumption growth.

(Between September 2016-February 2017)

Installment Expenditure Change (Percent) 15

Installment Cap Change (Months, RHE)

12 9 6

5 0 -5

3 0 -3 -6

-10 -15

-9 -12

It seems that the credit card installment cap regulation has had the expected impact on installment expenditures. While credit card expenditures with installments increased after September 2016 for sectors where the installment limit was raised or kept unchanged, they decreased for other sectors in which the installment limit was lowered. In February 2017, 55 percent of credit card expenditures

Note: Individual credit card makes up about 80 percent of total credit card balance. Annualized change of expenditure represents yearly change of expenditures. Positive (negative) change on installment cap represents increase (decrease) in the limit.

Source: BKM

with installments stemmed from sectors that saw their

installment

limits increased by the latest regulation. Furniture, construction materials and insurance sectors contributed significantly to the increase in credit card expenditures with installments. Although the

Financial Stability Report â&#x20AC;&#x201C; May 2017

Central Bank of the Republic of Turkey TĂźrkiyr installment limit for credit card expenditures with installments was increased by 3 months in the services sector, no significant increase was observed in this sectorâ&#x20AC;&#x2122;s expenditures with installments. On the other hand, installment purchases dropped significantly in the

Chart IV.4.5 Credit Card Installment Transactions in Furniture and Decoration Sectors (Million TL) Installment Transaction Amount

cosmetics sector following the reduction in the installment cap

VAT Deduction

350

(Chart IV.4.4).

325 300 275

250

purchases and the value added tax rate cut in furniture purchases in

225

early February 2017 had positive effects on sales with credit cards. In

175 04.17

03.17

02.17

01.17

12.16

11.16

10.16

09.16

08.16

07.16

06.16

05.16

expenditures with installments in the weeks following the regulation

150 03.16

the furniture sector, there has been a slight revival in credit card

200

04.16

The abolition of the special consumption tax on white goods

Source: CBRT (Latest Data: 28.04.17)

(Chart IV.4.5). Chart IV.4.6

IV.4.2

The Restructuring Facility for General Purpose

Restructured Retail Loans Monitored Under Performing Loans

Loans and Individual Credit Card Debt In September 2016, a restructuring facility was introduced to

enable restructuring with a maximum maturity of 72 months for ICC

debt and general purpose loans extended before that date. The

restructuring in retail loans increased in the last quarter of 2016. The

increase in the restructuring largely originated from standard loans,

which indicates that the restructuring was mostly intended to

Standard (Billion TL) Close Monitoring (Billion TL) Standard (3-Month % Change) Close Monitoring (3-Month % Change)

48 38

12 28

18 8

06.16

09.16

12.16

03.17

-2 -12

lengthen loan maturity to ease the debt service rather than

Note: Data covers 14 banks that represent 97 percent of retail loans.

payment difficulties (Chart IV.4.6).

Source: Audit Reports

Although the restructuring is not expected to have

impact on the stock of general purpose loans, it has already had a

Chart IV.4.7 General Purpose Loan Extensions and Interest Rates (4 Weeks Ave., Flow, Billion TL, Percent)

significant influence on flow extensions. Thanks to the moderate course of financial conditions, general purpose loan extensions

Extension (Billion TL) Interest Rate (Percent, RHE)

4,5

extended by increasing the cap on the ratio of loan amount to the

04.17

12.16

08.16

04.16

12.15

Loan financing facilities for house purchases have been

08.15

0,0

04.15

0,5

Housing Loans

12.14

1,0

Raising the Cap on Loan to Value Ratio in

08.14

1,5

IV.4.3

04.14

2,0

12.13

2,5

08.13

3,0

04.13

IV.4.7).

12.12

3,5

08.12

showed a significant improvement in the last quarter of 2016 (Chart

4,0

Source: CBRT (Latest Data: 28.04.17)

value of the property taken as collateral (LTV) from 75 to 80 percent. The fact that the actual LTV is at 55-60 percent, well below the cap of 75 percent set in 2010, implies that the impact of the change in the LTV cap on the sector may be limited. However, it would not be

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr a realistic approach to assume that the credit demand for housing

Chart IV.4.8 Loan to Value Ratio for Housing Loans

financing is homogeneous. Some customers make higher down

(Percent)

payments while others may prefer or need to receive a loan at the

Actual Loan to Value Ratio Loan to Value Cap Interest Rate (Other costs included, RHA) 16

legal cap. Therefore, all other things being equal, raising the LTV cap

can increase the actual LTV which can be regarded as the credit

12 74

financing tendency. In fact, the actual LTV has increased in the

recent period in which the new regulation has been in force. In this

period, interest rates declining significantly have also increased the

4 59

effect of the regulation via their essential role in the rise in the actual

04.17

01.17

10.16

07.16

04.16

01.16

10.15

07.15

04.15

01.15

10.14

07.14

04.14

01.14

LTV (Chart IV.4.8).

Source: CBRT

Housing loan interest rate cuts, which started with some large real estate companies in August 2016 and were followed by other

Chart IV.4.9 Contributions to Annual Change in Housing Sales

companies and banks, and the VAT cut in housing sales in

(Percent)

Second Hand Sales First Sales Total Change Interest Rate (Other Cost Included, RHA) 55 35

September 2016 led to a more positive outlook for the housing 16

market in the second half of 2016 compared to the first half of 2016

(Chart IV.4.9). Campaigns supported the housing market through the

demand channel by making house purchases more attractive and

housing sales have performed better than the first half of 2016,

especially with the contribution of mortgage sales. The moderate

trend in housing loan interest rates in the second half of 2016 has

15 -5

01.17

10.16

07.16

04.16

01.16

10.15

07.15

04.15

01.15

10.14

07.14

04.14

01.14

-25

facilitated an increase in banks’ extension of housing loans.

Source: CBRT (Latest Data: 03.17)

IV.4.4

Chart IV.4.10

Reduction in General Provisions

Housing Loan Extensions and Interest Rates (4 Week-Average, Flow)

The practice of allocating a general provision of four times the

Amount Extended (Billion TL) Interest Rate (Percent, RHE)

4,5

the retail loan NPL rates and loan compositions of banks, was abandoned in September 2016. General provisions were set at 1 percent for standard consumer loans and at 2 percent for closely monitored

consumer

loans,

regardless

the

banks’

loan

compositions and NPL ratios (Table IV.4.3). In practice, banks have

04.17

0,0

12.16

08.16

0,5

04.16

12.15

1,0

08.15

1,5

04.15

12.14

2,0

08.14

2,5

04.14

12.13

3,0

08.13

3,5

04.13

12.12

4,0

08.12

provision for consumer loans excluding housing loans, depending on

maintained a prudent provision policy although general provisions

Source: CBRT (Latest Data: 28.04.17)

were reduced.

Table IV.4.3 General Provision Ratios for Consumer Loans (Percent)

Consumer Loans Standard

Under Close Monitoring

Consumer Loans Except Vehicle and Housing* Under Close Standard Monitoring

Unstructured

Restructured

27 September 2016 and after

After 1 January 2018 (Banks Not Practicing TFRS 9)

1,5

Before 27 September 2016

* For banks whose share of consumer loans other than housing loans in total loans exceeds 25 percent, or whose NPL ratio for consumer loans other than houising loans exceeds 8 percent.

Source: CBRT

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey TĂźrkiyr

IV.4.5

Conclusion

Comprehensive steps, including macroprudential regulations,

Chart IV.4.11 Housing and General Purpose Loans (Indexed, September 2016=100)

have supported economic growth and contributed positively to the

Housing 27.09.2016

asset quality outlook of banks since the last quarter of 2016. After

115

September 2016, the retail loan growth has strengthened on the

110

back of housing and general purpose loans (Chart IV.4.11).

105

Regulations regarding the general purpose loan maturities and ICC

100

opportunity to ease their debt service by lengthening their debt maturity. On the other hand, the moderate course of housing loan interest rates and the increase in the LTV cap have increased the

95 90 -31 -28 -25 -22 -19 -16 -13 -10 -7 -4 -1 2 5 8 11 14 17 20 23 26 29

installments are believed to have provided customers with the

General Purpose

Week Source: CBRT (Latest Data: 28.04.17)

tendency for loan financing in housing purchases. With the contribution of these developments, housing loans have become one of the major drivers of retail loan growth.

Financial Stability Report - May 2017

Central Bank of the Republic of Turkey Türkiyr

IV.5

The Role of Bank Characteristics in the Interest Rate Transmission

Summary

In this note, we study how transmission of policy rates on corporate lending rates differs across banks. We control firm characteristics and demand side effects, and focus on the supply side. The results suggest that strongly capitalized or relatively liquid banks reflect changes in the policy rate less onto their lending rates, which overlaps with the literature for advanced economies. Moreover, we observe that banks with higher non-core foreign currency liabilities reflect policy changes less strongly. In this respect, we analyze the potential role of global liquidity cycles in the policy rate transmission.

IV.5.1

Introduction

Changes in monetary policy rate transmit to corporate lending rates directly by affecting banks’ average cost of funding or indirectly by signaling future policy actions or economic activity. Changes in lending rates, in turn, affect macroeconomic

aggregates

such

overall

demand

conditions, economic activity and inflation. In this respect, it is important to quantify how much lending rates respond to changes in the policy rate. By definition, equilibrium corporate lending rates are set depending on demand-(firm) or supply-(bank) characteristics. Not only firm characteristics (such as firm size, leverage, collateral ratio, value of relationship with a bank, cash need) but also bank characteristics and banks’ supply behavior have a direct bearing on the lending rates. Moreover, macroeconomic aggregates such as overall economic activity and inflation can affect lending rates as well. Considering that monetary policy can affect firm or bank balance sheets and macroeconomic conditions, the observed pass-through of policy rates onto lending rates entails many factors. In this regard, isolating how changes in policy rates affect banks’ setting their lending rates (independent from indirect effects such as the effect of policy rates on firm balance sheets or aggregate economic conditions)

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey Türkiyr is challenging. For a strict identification, two observations are in order: (i)

Monetary policy affects banks’ supply behavior at different degrees. For instance, banks reflect changes

(ii)

Graph IV.5.1

in monetary policy rates onto their loan portfolios

Global Liquidity Conditions and Banks’ Non-core Foreign Currency Liabilities

differently depending on their capital, liquidity or

access to external funds.

Focusing on similar firms helps filter out monetary

policy’s indirect effects on firm balance sheets, or

demand-side effects.1 For instance, consider two

-10

banks --that differ only in terms of their capital ratios--

-30

(Annual Percentage Change, US dollar)

-30 1234123412341234123412341234 2010 2011 2012 2013 2014 2015 2016 Fed Total Assets

lending to the same firm. Following a monetary policy

FX Non-Core Liabilities

tightening, these banks may set different lending rates to the same firm, the difference reflecting the role of

Source: CBRT, Federal Reserve.

bank capital in the interest rate transmission. Our empirical strategy rests on (i) and (ii). To this end, we study the effect of changes in monetary policy rate (namely, the weighted average cost of funding)2 on how banks set their lending rates. Note also that non-core foreign currency liability of banks operating in Turkey moves in tandem with global liquidity cycles (Graph IV.5.1 and Graph IV.5.2). Along these

Graph IV.5.2 Global Liquidity Conditions and Banks’ Non-core Foreign Currency Liabilities (Annual Percentage Change, US dollar)

40 30

lines, we focus not only on bank capital or liquidity as potential

determinants of how the transmission differs across banks, but

also non-core foreign currency liabilities. We, therefore, quantify in a well-identified way whether global liquidity cycles matter for

-5

-10

-20 -30

-15 1234123412341234123412341234123412341234123412341234 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016

interest rate transmission. FX Non-Core Liab.(in USD, y-o-y % change)

The results suggest that strongly capitalized or relatively liquid banks raise their lending rates less following a monetary

Global Liquidity (International Bank to Bank Claims, y-o-y % change, right axis) Source: CBRT, BIS.

policy tightening. Similarly, banks with higher non-core foreign currency liabilities reflect policy tightening less onto their lending rates. The results are statistically significant and economically relevant. Following a 100 basis points-increase in the policy rate, strongly capitalized banks (the bank at the 75 th percentile of capital ratio compared to the 25th percentile) raise their lending rates by 32 basis points less. Similarly, relatively liquid banks or

1 For example, monetary policy can affect asset prices, and in turn, the value of collateral pledged by firms. Similarly, monetary policy can affect exchange rates, and thus, importing firms’ production costs or leverage. 2 The Central Bank of the Republic of Turkey has implemented a multiple interest rate framework to respond, if needed, to market volatility in a timely manner, starting in late 2010. For details on the framework, see Binici, Kara and Özlü (2016).

Financial Stability Report – May 2017

Central Bank of the Republic of Turkey TĂźrkiyr banks with high non-core liabilities raise their lending rates by 25 and 50 basis points less, respectively. It has been well established that banks may react to changes in monetary policy at different strengths. For instance, Kashyap and Stein (2000) show for the US that less liquid or small banks reflect changes in Fed policy more strongly on their supply of credit. Using a stronger identification scheme, Jimenez et al. (2012) and Iyer et al. (2014) show, for Spain and Portugal, respectively that banks reflect market liquidity shocks differently depending on their capital and liquidity stance. Shedding further light on the transmission, we show whether exposure to global liquidity may matter for the transmission. As such, the main contribution of this analysis is to show in a well identified way the effect of global liquidity on the interest rate transmission. The note proceeds as follows: Section 2 presents the data set and the methodology, Section 3 the results and Section 4 concludes.

IV.5.2

Data Set and Methodology

Two different data sets are used in this study. These are data from the CBRT that include loan information at the firmbank level and the data from the Banking Regulation and Supervision Agency (BRSA) that cover banks' balance sheets and income statements. Both data sets are at a monthly frequency. Loan information at the firm-bank level is matched with bank data. Firm-bank level data set also includes information on the loan type (for which purpose the loan is demanded, maturity etc.). For econometric analysis, loan data set is collapsed at the bank-firm-loan type level. In the study, the following model is used to analyze how transmission of policy rates on corporate lending rates differs according to bankâ&#x20AC;&#x2122;s capital, liquidity and funding structure: 3

đ?&#x2018;&#x2013;b,f,a,t = â&#x2C6;&#x2018; đ?&#x203A;˝1,đ?&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; + â&#x2C6;&#x2018; đ?&#x203A;˝2,đ?&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; â&#x2C6;&#x2014; đ?&#x2018;&#x2039;đ?&#x2018;?,đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1

đ?&#x2018; =1

+đ??śđ?&#x2018;&#x153;đ?&#x2018;&#x203A;đ?&#x2018;Ąđ?&#x2018;&#x;đ?&#x2018;&#x153;đ?&#x2018;&#x2122;đ?&#x2018; +Îľb,f,a,t đ?&#x2018;&#x2013;b,f,a,t , stands for the loan rate on the credit provided by bank b at month t to firm f with loan type a. đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ą , is the monthly change

Financial Stability Report â&#x20AC;&#x201C; May 2017

Central Bank of the Republic of Turkey TĂźrkiyr in the weighted average funding cost of the CBRT funding which is taken as the monetary policy rate. One-to-three month lags (one quarter lag) of monetary policy rate is included considering that banks take into account previous funding costs while determining the loan rate. đ?&#x2018;&#x2039;đ?&#x2018;?,đ?&#x2018;Ą , denotes the bank variables on capital, liquidity and funding structure of bank b. Specifically, banksâ&#x20AC;&#x2122; capital adequacy ratio (đ??śđ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;), liquidity ratio (đ??żđ?&#x2018;&#x2013;đ?&#x2018;&#x17E;đ?&#x2018;˘đ?&#x2018;&#x2013;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ś đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;)

or the ratio of non-core foreign currency

liabilities to total assets (đ?&#x2018; đ?&#x2018;&#x153;đ?&#x2018;&#x203A; â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;&#x2019; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;) are used. These bank variables are included in levels as well as interactions with monetary policy rate to reveal how banks with different capital, liquidity and non-core funding ratios transmit changes in monetary policy rate to corporate loan rates. đ??śđ?&#x2018;&#x153;đ?&#x2018;&#x203A;đ?&#x2018;Ąđ?&#x2018;&#x;đ?&#x2018;&#x153;đ?&#x2018;&#x2122;đ?&#x2018; include bank balance sheet characteristics, macroeconomic indicators, variable showing strength of bankfirm relationship, and a large set of fixed effects. Bank controls are: capital adequacy ratio (đ??śđ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;), total assets, liquidity ratio (đ??żđ?&#x2018;&#x2013;đ?&#x2018;&#x17E;đ?&#x2018;˘đ?&#x2018;&#x2013;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ś đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;), non-performing loans ratio, return on assets, the ratio of non-core foreign currency liabilities to total assets (đ?&#x2018; đ?&#x2018;&#x153;đ?&#x2018;&#x203A; â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;&#x2019; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;) and lastly Herfindahl by bank (bankâ&#x20AC;&#x2122;s share of total banking loans to a particular sector). All bank controls are lagged by one month. When interactions of (đ??śđ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;), (đ??żđ?&#x2018;&#x2013;đ?&#x2018;&#x17E;đ?&#x2018;˘đ?&#x2018;&#x2013;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ś đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;), and (đ?&#x2018; đ?&#x2018;&#x153;đ?&#x2018;&#x203A; â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;&#x2019; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;) with monetary policy rate are included in the model, one-to-three month lags of the corresponding variable and interactions with the policy rate are included as well.3

3 For instance, in the specification that analyses the impact of bank capital on interest rate transmission, where đ?&#x2018;&#x2039;đ?&#x2018;?,đ?&#x2018;Ą is capital ratio, control variables are one-month lags of total assets, liquidity ratio (đ??żđ?&#x2018;&#x2013;đ?&#x2018;&#x17E;đ?&#x2018;˘đ?&#x2018;&#x2013;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ś đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;), non-performing loans ratio, return on assets, the ratio of non-core foreign currency liabilities to total assets (đ?&#x2018; đ?&#x2018;&#x153;đ?&#x2018;&#x203A; â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;&#x2019; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;) and Herfindahl by bank (bankâ&#x20AC;&#x2122;s share of total banking loans to a particular sector) and one-, two- and three- month lags of capital ratio.

Financial Stability Report â&#x20AC;&#x201C; May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Table IV.5.1 Summary Statistics (January 2005 â&#x20AC;&#x201C; December 2016)

Unit

No. of observations

Mean

Minimum

Median

Maximum

St. Dev

Bank-Firm Level Variables Loan Rate Strength of Bank-Firm Relationship

8,190,595

11.52

0.51

10.732

99.84

8.239

[0,1]

8,190,595

0.287

0.205

0.249

6,332

25.996

1.618

14.428

99.952

24.035

14.965

7.915

15.053

19.677

2.395

0.267

32.188

99.699

25.009 11.030

Bank Level Variables Capital Ratio

000 TL (Log)

Total Assets (Log)

6,332

Liquidity Ratio

6,332

41.566

Non-Performing Loans Ratio

6,031

2.138

0.473

363.226

Return on Assets

6,332

1.235

-19.376

1.266

20.751

2.323

Non-Core FX Liabiltiies Ratio

6,332

21.214

15.064

94.822

20.624

Herfindahl by bank

55,667

4.504

1.343

100

8.440

Î&#x201D; MP (Monthly)

144

-0.084

-1.845

-0.015

2.567

0.579

Î&#x201D; IPI (Annual)

131

3.796

-19.998

4.354

18.199

6.997

Î&#x201D; CPI (Annual)

144

8.259

3.986

8.169

12.065

1.637

Î&#x201D; EMBIG Turkey (Monthly)

144

0.004

-1.150

-0.020

2.450

0.344

Macroeconomic Variables

As for macroeconomic indicators, annual growth in industrial production index (IPI), annual percentage change in

Table IV.5.2 Empirical Results

consumer price index (CPI), and the monthly change in the

Dependent Variable: Loan Rate at bank-firm-loan level (percentage point)

(1)

(2)

(3)

0.526***

0.992***

(0.035)

(0.039)

-0.06***

-0.079***

(0.002)

Yes No

Yes Yes

Yes -

Yes

Yes Yes Yes Yes No 8,189,665 0.428

Yes Yes Yes Yes No 8,108,822 0.447

Yes Yes No Yes 7,828,395 0.579

â&#x2C6;&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1

the

â&#x2C6;&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; â&#x2C6;&#x2014; đ?&#x2018; đ?&#x2018;&#x153;đ?&#x2018;&#x203A; â&#x2C6;&#x2019; đ?&#x2018;?đ?&#x2018;&#x153;đ?&#x2018;&#x;đ?&#x2018;&#x2019; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;.đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1

Bank Controls Macroeconomic Variables Interaction (Macroeconomic v. with Non-Core Ratio) Strength of Bank-Firm Relationship Firm Fixed Effect Bank Fixed Effect Loan Type Fixed Effect Firm-Time Fixed Effect No. of observations R2 Following 100 basis points increase in the policy rate Difference in the lending rate (basis points) (Banks with high non-core ratio compared to banks with low noncore ratio (p75-p25))

EMBIG-Turkey are included. In line with the lag specification for monetary

policy

rate,

one-to-three

month

lags

macroeconomic variables are included in the estimation. The strength of bank-firm credit relationship is measured by the ratio of credits a firm obtains from a particular bank in the last 12 months to the total credit the firm obtains from all banks during the same period. All specifications include firm and bank fixed effects. In some specifications, loan type and firm-time fixed effects are included as well. Descriptive statistics of the variables used in the empirical analyses are provided in Table IV.5.1.

-21.8

-18.8

-51.8

Note: The results are obtained using ordinary least squares. Sample period: 2005:12016:12. For identification, the sample is restricted to firms that work with at least two banks. All control variables are lagged by one month. Regarding the fixed effects, "Yes" indicates that corresponding fixed effects (or the variable) are included. "No" indicates that corresponding fixed effects (or the variable) are not included. "--" indicates that the respective fixed effect is inapplicable or already included in the wider set of fixed effects or variables. Robust standard errors are in parentheses. *** Significant at 1%, ** significant at 5%, and * significant at 10%.

IV.5.3

Empirical Results

Table IV.5.2 presents the main results. Column (1) presents the most parsimonious specification. As control variables, bank characteristics and the strength of bank-firm relationship are used.4

4 Strength of bank-firm relationship is defined as total loans a firm obtains from a particular bank during the last 12 months compared to total loans the firm obtains from all banks during the same period.

Financial Stability Report â&#x20AC;&#x201C; May 2017

Central Bank of the Republic of Turkey TĂźrkiyr Column (2) further includes macroeconomics variables (in levels and in interactions) as controls. Column (3), the most general specification, includes firm-month fixed effects (thereby control for demand side effects). In all the columns, we observe that banks with high non-core ratio (with non-core ratio at the 75th percentile) raise their lending rates less (by about 50 basis

Table IV.5.3 Empirical Results

Dependent Variable: Loan Rate at bank-firm-loan level (percentage point)

â&#x2C6;&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1

â&#x2C6;&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; â&#x2C6;&#x2014; đ??śđ?&#x2018;&#x17D;đ?&#x2018;?đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;&#x17D;đ?&#x2018;&#x2122; đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1

tightening less onto their loan rates (Table IV.5.3 and Table IV.5.4). Based on the last column of Table IV.5.3 (which is the most

Following 100 basis points increase in the policy rate

core ratio at the 25th percentile). Moreover, in line with the literature, we observe that strongly capitalized or relatively liquid banks reflect policy

general empirical specification), strongly capitalized banks (with capital ratio at the 75 percentile) raise their lending rates by 32 basis points less compared to weakly capitalized banks (with capital ratio at the 25th percentile). Similarly, relatively liquid banks raise their lending rates by 25 basis points less (Table IV.5.4).

IV.5.4

(2)

(3)

0.108***

0.326***

(0.041)

(0.046)

-0.043***

-0.055***

-0.05***

(0.004)

Yes No

Yes Yes

Yes -

Bank Controls Macroeconomic Variables Interaction (Macroeconomic v. with Capital Ratio) Strength of Bank-Firm Relationship Firm Fixed Effect Bank Fixed Effect Loan Type Fixed Effect Firm-Time Fixed Effect No. of observations R2

points) compared to banks with low non-core ratio (with non-

(1) 3

Difference in the lending rate (basis points) (Banks with high capital ratio compared to banks with low capital ratio (p75-p25))

Yes

Yes Yes Yes No 8,189,665 0.427

Yes Yes Yes No 8,108,822 0.446

Yes No Yes 7,828,395 0.579

-24.1

-24.8

-32.2

Conclusion

In this note, we study the transmission of policy rates onto corporate lending rates for 2005-2016 period. For better identification, we focus on the heterogeneity among banks in their responses to changes in the monetary policy rate. Results suggest that, banks have access to internal or external funds â&#x20AC;&#x201C; i.e. strongly capitalized or relatively liquid banks, or banks with higher non-core foreign currency fundsâ&#x20AC;&#x201D;raise their lending rates less following a monetary policy tightening.

It should be noted that using the surprise component of policy rate changes would give a sharper picture. Still, we would like to highlight that our exhaustive set of control variables, including macroeconomic variables in levels and in interaction with bank characteristics, are likely to make our results by and large robust. We leave the effect of changes in the policy rate on credit volume and the effect of accommodative fiscal policies on the interest transmission to future research.

Financial Stability Report â&#x20AC;&#x201C; May 2017

Table IV.5.4 Empirical Results

Dependent Variable: Loan Rate at bank-firm-loan level (percentage point)

(1)

(2)

0.051***

0.591***

(3)

(0.048)

(0.055)

-0.017***

-0.032***

-0.022***

(0.002)

Yes No

Yes Yes

Yes -

Yes

Yes Yes Yes No 8,189,665 0.427

Yes Yes Yes No 8,108,822 0.446

Yes No Yes 7,828,395 0.579

-17.7

-16.7

-25.1

â&#x2C6;&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1 3

â&#x2C6;&#x2018; đ?&#x203A;Ľđ?&#x2018;&#x20AC;đ?&#x2018;&#x192;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; â&#x2C6;&#x2014; đ??żđ?&#x2018;&#x2013;đ?&#x2018;&#x17E;đ?&#x2018;˘đ?&#x2018;&#x2013;đ?&#x2018;&#x2018;đ?&#x2018;&#x2013;đ?&#x2018;Ąđ?&#x2018;Ś đ?&#x2018;&#x2026;đ?&#x2018;&#x17D;đ?&#x2018;Ąđ?&#x2018;&#x2013;đ?&#x2018;&#x153;đ?&#x2018;Ąâ&#x2C6;&#x2019;đ?&#x2018; đ?&#x2018; =1

Bank Controls Macroeconomic Variables Interaction (Macroeconomic v. with Liquidity Ratio) Strength of Bank-Firm Relationship Firm Fixed Effect Bank Fixed Effect Loan Type Fixed Effect Firm-Time Fixed Effect No. of observations R2 Following 100 basis points increase in the policy rate Difference in the lending rate (basis points) (Banks with high liquidity ratio compared to banks with low liquidity ratio (p75-p25))

Central Bank of the Republic of Turkey Türkiyr

References

Binici, M., Kara, H., and P. Özlü (2016). Faiz Koridoru ve Banka Faizleri: Parasal

Aktarım Mekanizmasına Dair Bazı

Bulgular. CBRT Working Paper No. 16/08. (available in Turkish only)

Iyer, R., J.-L. Peydro, S. da Rocha-Lopes, and A. Schoar (2014). Interbank Liquidity Crunch and the Firm Credit Crunch: Evidence from the 2007-2009 Crisis. Review of Financial Studies 27, 347–372.

Jimenez, G., S. Ongena, J. Peydro, and J. Saurina (2014). Credit supply and monetary policy: Identifying the bank balance-sheet

channel

with

loan

applications.

American Economic Review 102, 2301–2326.

Kashyap, A. K. and J. C. Stein (2000). What do a million observations on banks say about the transmission of monetary policy. American Economic Review 90(3), 407–428.

Financial Stability Report – May 2017

Appendix Appendix 1 - Macroprudential Regulations on Retail and Corporate Loans Macroprudential Tools

Regulation

Macroprudential Measures (Tightening)

Enforcement Date

I ncreasing minimum payment rat es of credit cards

17.12.2010 01.01.2014 (kademeli

Macroprudential Measures (Easing)

Enforcement Date

geçiş)1 Regulat ion on Bank Cards and Credit Cards

Credit restrictions

Regulat ion on Credit Operat ions of Banks Regulat ion on Principles for Est ablishment and Operat ions of Financial Leasing, Fact oring and Financing Companies

Capital adequacy

Regulat ion on Measurement and Assessment of Capit al Adequacy of Banks

Loan/I ncome rat io rest rict ion for credit card limit s

08.10.2013

Closing credit cards in cert ain cases

17.06.2011 08.10.2013

Limit ing t he inst allment period of credit card debt

01.02.2014 13.05.2014 22.10.2014

Ext ending t he inst allment period of credit card debt

25.11.2015 27.09.2016

Rest rict ing mat urit y of vehicle loans t o 48 mont hs and consumer loans except housing and vehicle loans t o 36 mont hs

31.12.2013

Ext ending mat urit y rest rict ion for consumer loans except housing and vehicle loans t o 48 mont hs

27.09.2016

Loan/Value rat io rest rict ion (75%) for housing loans ext ended t o consumers

01.01.2011

I ncreasing t he Loan/Value rat io rest rict ion t o 80% for housing loans ext ended t o consumers

27.09.2016

Loan/Value rat io rest rict ion (50%) for housing loans ext ended for commercial purposes

01.01.2011

Abolit ion of Loan/Value rat io rest rict ion for housing loans ext ended for commercial purposes

04.04.2013

Loan/Value rat io rest rict ion for vehicle loans (70% up t o 50,000 TL of t he billing value, 50% for t he rest )

01.02.2014

I ncreasing risk weight s applied t o consumer loans and credit cards

22.03.2008 18.06.2011 08.10.2013

Decreasing risk weight s applied t o consumer loans and credit cards (To comply wit h Basel regulat ions)

01.07.2012 31.03.2016

18.06.2011 08.10.2013

Abolit ion of increment al general provision rat es for consumer loans except housing loans

27.09.2016

I ncrement al general provision rat es for consumer loans except housing

General provisions

loans

Regulat ion on Procedures and Principles for Det erminat ion of Qualificat ions of Loans and Ot her Receivables by Banks and Provisions t o Be Set Aside

24.12.2013

Reducing general provision rat es • t o 0% for export loans4, • by half for SME loans, which are cat egorized in t he first group

Regulat ion on Account ing Applicat ions and Financial Tables of Financial Leasing, Fact oring and Financing Companies

Reserve requirements

Communiqué on Reserve Requirement s

I nclusion of financing companies in t he reserve requirement syst em

08.10.2013

Reducing general provision rat es • t o 0% for SME loans and syndicat ed loans for large-scale public procurement s, • by half for commercial loans, which are cat egorized in t he first group

14.12.2016

Reducing general provision rat es by half for commercial loans, SME loans and export loans, which are cat egorized in t he second group

14.12.2016

06.12.2013

The minimum payment rat es for credit cards were gradually increased every 6 mont hs from t he end of 2010 t ill t he end of 2013, at t he beginning of 2014 and at t he beginning of 2015. The t imet able for t he gradual t ransit ion t o minimum payment rat es is given on page 52 of t he Financial St abilit y Report November 2014, I ssue 19. 2

The applicat ion is det ailed in pages 57 and 58 of t he Financial St abilit y Report November 2014, I ssue 19.

The enforcement dat e of t he regulat ion for finance companies.

The t ransact ions t hat are deemed as export loans have been clarified wit h t he regulat ion t hat ent ered int o force on 14.02.2015, so t he loans ext ended t o t ransit commercial, export sales and deliveries and foreign exchange earning services and act ivit ies are evaluat ed in t his scope.

Charts, Tables and Figures Overview Chart 1

Weekly Capital Flows to Emerging Markets ....................................................................................... i

Chart 2

Exchange Rate Indices ......................................................................................................................... i

Chart 3

Annual Loan Growth ............................................................................................................................. ii

Chart 4

Loan Rates and Regulations ................................................................................................................ ii

Chart 5

Ratio of Non-Deposit Funding to Funding Sources............................................................................ iii

Chart 6

Loan/Deposit Ratio ................................................................................................................................ iii

Chart 7

NPL Ratios................................................................................................................................................ iv

Chart 8

Corporate NPL Ratios ............................................................................................................................ iv

Chart 9

Return on Assets (ROA) and Return on Equities (ROE) ...................................................................... iv

Chart 10

Spread between Loan and Deposit Rates......................................................................................... iv

Chart 11

CAR and Core Tier 1 CAR ..................................................................................................................... v

Chart 12

Changes in Items Affecting Capital .................................................................................................... v

I. Macroeconomic Outlook I.1. International Developments Chart I.1.1

Economic Policy Uncertainty Indices .................................................................................................. 2

Chart I.1.2

FOMC Members' Median Policy Interest Forecasts and Market Expectations.............................. 2

Chart I.1.3

Stock Market Indices ............................................................................................................................. 2

Chart I.1.4

10-Year Treasury Bond Yields in US, Germany and Japan ............................................................... 3

Chart I.1.5

10-Year Treasury Bond Rate in Emerging Economies ....................................................................... 3

Chart I.1.6

Weekly Capital Flows to Emerging Economies .................................................................................. 3

Chart I.1.7

CDS Premiums in Emerging Economies .............................................................................................. 3

Chart I.1.8

Exchange Rate Indices ......................................................................................................................... 4

Chart I.1.9

Growth of Advanced and Emerging Economies ............................................................................. 4

Chart I.1.10

Manufacturing Industry PMI Indices .................................................................................................... 4

Chart I.1.11

Commodity Prices ................................................................................................................................. 5

Chart I.1.I.1

US Trade Deficit Composition ............................................................................................................... 7

Table I.1.I.1

Turkeyâ&#x20AC;&#x2122;s Exports to the US ...................................................................................................................... 7

Chart I.1.I.2

Weighted Average Effective Tariff Rates Applied by the US .......................................................... 8

Table I.1.II.1

Conditions for Impairment and Stages ............................................................................................... 10

Table I.1.II.2

Example: Calculation of Expected Credit Loss ................................................................................. 11

Chart I.1.II.1

Provisions for Banks Conforming to the TFRS 9 ................................................................................... 12

Chart I.1.II.2

Provisions for Banks Not Conforming to the TFRS 9 ............................................................................ 12

I.2. Domestic Developments Chart I.2.1

Contribution to Growth from the Expenditure Side........................................................................... 13

Chart I.2.2

Industrial Production Index ................................................................................................................... 13

Chart I.2.3

Labor Force ............................................................................................................................................ 13

Chart I.2.4

Price Indices ........................................................................................................................................... 14

Chart I.2.5

Current Account .................................................................................................................................... 14

Chart I.2.6

Foreign Trade ......................................................................................................................................... 14

Chart I.2.7

Current Account Deficit Financing Items ........................................................................................... 15

Chart I.2.8

Central Government Budget Balance ............................................................................................... 15

Chart I.2.9

Exchange Rate Basket and CDS ......................................................................................................... 15

Chart I.2.10

Interest Rates .......................................................................................................................................... 15

II. Non-Financial Sector II.1. Household Development Chart II.1.1

Household Financial Assets’ and Liabilities’ Growth Rates and Financial Leverage Ratio .......... 18

Chart II.1.2

Household Loans and Deposits Growth .............................................................................................. 18

Chart II.1.3

Savings Deposits and Consumer Loans Interest Rates ...................................................................... 18

Table II.1.1

Household Financial Assets ................................................................................................................... 18

Chart II.1.4

Savings Deposits of Resident Households By TL and FX Breakdown ................................................ 19

Chart II.1.5

Contribution of Resident Households’ Deposit Amounts to Growth by Periods ............................ 19

Chart II.1.6

Households’ Gold Portfolio in the Banking System and Gold Prices ............................................... 19

Chart II.1.7

Private Pension System in Turkey .......................................................................................................... 20

Chart II.1.8

Automatic Enrollment to Private Pension System .............................................................................. 20

Chart II.1.9

BİST All Index and Household Equity Securities Portfolio ................................................................... 20

Table II.1.2

Household Financial Liabilities .............................................................................................................. 21

Chart II.1.10

Consumer Loans Extended by Financing Companies Based on Type.......................................... 21

Chart II.1.11

Average Retail Loan Maturity .............................................................................................................. 22

Chart II.1.12

Contribution to Housing Sales Growth, Housing Loan Monthly Interest Rate and Granted Loan Ratio ................................................................................ 22

Chart II.1.13

Individual Credit Card Balance ........................................................................................................... 22

Chart II.1.I.1

Usage of ROM for Gold ....................................................................................................................... 24

Chart II.1.I.2

Usage of Gold Facility for Precious Metals ......................................................................................... 24

Chart II.1.I.3

Usage of ROM for Scrap Gold ............................................................................................................. 24

II.2. Real Sector Developments Chart II.2.1

Industrial Production and Investment Tendency ............................................................................... 26

Chart II.2.2

Real Sector Confidence Index ............................................................................................................. 26

Chart II.2.3

Automobile and White Goods Production Volume Annual Growth and Seasonally Adjusted Export Quantity Index ........................................................................................ 26

Chart II.2.4

Share of Real Sector Financial Debt in GDP and Annual Growth of FX Loans .............................. 27

Chart II.2.5

International Comparison of Real Sector Credit / GDP Ratio .......................................................... 27

Chart II.2.6

Share of the Loan Debts of Main Sectors in the Sectoral Value Added and Total Loan Volume ................................................................................................ 27

Chart II.2.7

Share of Loans Granted for Project Financing in Total Loans .......................................................... 28

Chart II.2.8

Developments in SME Loans ................................................................................................................. 28

Chart II.2.9

TL Financial Costs of SMEs ..................................................................................................................... 28

Chart II.2.10

FX Open Position of Non-Financial Companies ................................................................................. 29

Chart II.2.11

Maturity Breakdown of Domestic FX Loans and External FX Liabilities ............................................ 29

Chart II.2.12

Distribution of FX Loans by Firm Number ............................................................................................. 29

Chart II.2.13

NPL Ratio in FX Loans ............................................................................................................................. 30

Chart II.2.14

Developments in FX Deposits of Domestic Firms ................................................................................ 30

Chart II.2.15

Financial Indicators of Publicly Listed Real Sector Companies ....................................................... 30

Chart II.2.16

Firms’ Profitability .................................................................................................................................... 31

Chart II.2.17

Sectoral Indebtedness and Interest Coverage Ratio ....................................................................... 31

Chart II.2.18

Sectoral FX Open Position / Assets ....................................................................................................... 32

Chart II.2.19

FX Open Position and Export Revenues .............................................................................................. 32

Chart II.2.I.1

Rediscount Credit Utilization and Outstanding Balance .................................................................. 34

Chart II.2.I.2

Contribution to CBRT Reserves ............................................................................................................. 34

III. Financial Sector III.1. Credit Developments and Credit Risk Chart III.1.1

Annual Loan Growth ............................................................................................................................. 36

Chart III.1.2

Credit/GDP Ratio ................................................................................................................................... 37

Chart III.1.3

Annual Change in Credit Stock to GDP ............................................................................................. 37

Chart III.1.4

International Comparison of Credit/GDP ........................................................................................... 37

Chart III.1.5

Annual Growth in TL Corporate Loans by Firm Size ........................................................................... 38

Chart III.1.6

Annual Growth in FX Corporate Loans by Size .................................................................................. 38

Chart III.1.7

Corporate Loan Interest Rates and Spreads ..................................................................................... 39

Chart III.1.8

Contributions to Corporate Loan Supply............................................................................................ 39

Chart III.1.9

Annual Growth in Retail Loans ............................................................................................................. 39

Chart III.1.10

Retail Loan Lending Rates .................................................................................................................... 40

Chart III.1.11

General Purpose Loan Weekly Growth Rates ................................................................................... 40

Chart III.1.12

General-Purpose Loan Maturities ........................................................................................................ 40

Chart III.1.13

Credit Standards and Economic Outlook .......................................................................................... 41

Chart III.1.14

NPL Ratios................................................................................................................................................ 41

Chart III.1.15

Components of NPL and Their Contributions to the Monthly Growth Rate of NPL ....................... 41

Chart III.1.16

Corporate NPL Ratios ............................................................................................................................ 42

Chart III.1.17

International Comparison of NPL Ratios and Differences ................................................................ 42

Table III.1.1

Sectoral Breakdown of NPL Ratios ...................................................................................................... 42

Chart III.1.18

NPL Ratios in Retail Loans...................................................................................................................... 43

Chart III.1.19

Growth in Personal Credit Card Balances and Installment Share .................................................. 43

Chart III.1.20

New General Purpose Loans and the Survey .................................................................................... 43

Chart III.1.21

General-Purpose Loan Maturities by RLS ............................................................................................ 44

Chart III.1.22

General-Purpose Loans Vintage Curves ............................................................................................ 44

III.2. Liquidity Risk Chart III.2.1

Quantiles of Banks by Total Liquidity Coverage Ratio ...................................................................... 45

Chart III.2.2

Quantiles of Banks by FX Liquidity Coverage Ratio .......................................................................... 46

Chart III.2.3

Ratio of Non-Deposit Funding to Funding Sources............................................................................ 46

Chart III.2.4

Loan/Deposit Ratio ................................................................................................................................ 47

Chart III.2.5

Loan/(Deposit+Other Stable Sources) Ratio ...................................................................................... 47

Chart III.2.6

Amount and Growth Rate of Banks’ External Liabilities .................................................................... 47

Chart III.2.7

Cost of Syndicated Loans with a Maturity of 367 days .................................................................... 48

Chart III.2.8

Change in Banks’ Short Term and Medium-Long Term External Liabilities ..................................... 48

Chart III.2.9

External Debt Roll-Over Ratio ............................................................................................................... 48

Chart III.2.10

External Debt Roll-Over Ratio and its Average Maturity .................................................................. 49

Chart III.2.11

Change in External Borrowing Instruments by the End of 2014 ....................................................... 49

Chart III.2.12

Change in External Borrowing Instruments by the End of 2014 ....................................................... 49

Chart III.2.13

ROM Reserves + FX Borrowing Facility and External FX Liabilities Due Within 1 Year .................... 50

Chart III.2.14

FX Issues Abroad .................................................................................................................................... 50

Chart III.2.15

Domestic TL Security Issues ................................................................................................................... 50

Figure III.2.I.1

Ranking Assets by Liquidity and Liabilities by Stability ....................................................................... 52

Chart III.2.I.1

Net Stable Funding Ratio in the Turkish Banking Sector .................................................................... 53

III.3. Interest Rate and Exchange Rate Risk Chart III.3.1

Short Term Open Position in TL .............................................................................................................. 54

Chart III.3.2

Short Term Open Position in FX ............................................................................................................. 54

Chart III.3.3

Interest Rate Risk via Repricing Channel Measured with Economic Value Approach ................ 54

Chart III.3.4

Interest Rate Risk on Securities with Fixed Interest Rate in Trading Portfolio.................................. 55

Chart III.3.5

FX position in the Banking Sector ......................................................................................................... 55

Chart III.3.6

Shares of Gross Positions of Off-Balance Sheet FX Transaction (Assets + Liabilities) ..................... 55

III.4. Profitability and Capital Adequacy Chart III.4.1

Return on Assets and Return on Equities (ROE) .................................................................................. 56

Chart III.4.2

CAR and Core Tier 1 CAR ..................................................................................................................... 56

Chart III.4.3

Effects of Income Statement Items on ROA ...................................................................................... 56

Chart III.4.4

Contribution to Changes in the Net Interest Income ....................................................................... 57

Chart III.4.5

Profit/Losses from Security, Derivative and FX Transactions ............................................................. 57

Chart III.4.6

Changes in Items Affecting Capital .................................................................................................... 57

Chart III.4.7

RWA Components ................................................................................................................................. 58

Chart III.4.8

CARs According to Bank Types ........................................................................................................... 58

Chart III.4.9

Banking Sector ‘s Capacity of Supporting Loan Growth under Current Profitability and Capital Adequacy Levels ........................................................................................ 58

Table III.4.I.1

Exchange Rate Effect on CAR Calculation ....................................................................................... 60

Chart III.4.I.1

Decomposition of the Effects on CAR ................................................................................................ 60

IV. Special Topics Chart IV.1.1

Number of Lender Countries ................................................................................................................ 63

Chart IV.1.2

Share of Lender Countries with the Highest Amount of Debt in Total Debt .................................. 63

Chart IV.1.3

Share of Lender Countries with the Highest Amount of Debt in Total Debt ................................... 63

Chart IV.1.4

Regional Distribution of External Debt ................................................................................................. 64

Chart IV.1.5

Regional Distribution of External Debt ................................................................................................. 64

Chart IV.1.6

Bank Indexes in Euro Zone and USA .................................................................................................... 64

Table IV.1.1

Global Liquidity Indicators .................................................................................................................... 65

Table IV.1.2

Estimation Results...................................................................................................................................... 67

Table IV.1.3

Estimation Results...................................................................................................................................... 69

Figure IV.2.1

Measures to Increase Corporate Access to Finance ....................................................................... 71

Figure IV.2.2

Compensation Upper Limit for Treasury-Supported KGF Guaranteed Loans ................................ 73

Chart IV.2.1

Corporate Loan Growth ....................................................................................................................... 75

Chart IV.2.2

Corporate Loan Growth By Scale ....................................................................................... 75

Chart IV.2.3

Corporate Loan Interest Rate by Scale .............................................................................................. 76

Chart IV.2.4

Loan Usage and Funding Structure between December 2016 – April 2017 .................................. 76

Chart IV.2.5

Deposit and 3 Month Currency Swap Interest Rates ........................................................................ 76

Table IV.3.1

Independent Variables ......................................................................................................................... 80

Chart IV.3.1

Average Firms’ Export Share (Annual) and Average FC Credit Share (Monthly) ......................... 81

Chart IV.3.2

Amount Distribution FC Credit Balance, as of December 2015 ...................................................... 81

Chart IV.3.3

Average Bank FC Liabilities Share and Average FC Credit Share .................................................. 82

Chart IV.3.4

Shares of FC Liabilities in Banks’ Total Liabilities and Average Firm FC Credits Share ................... 82

Table IV.3.2

Summary Statistics .................................................................................................................................. 82

Table IV.3.3

Coefficient Estimates, Dependent Variable: FC Credits/Total Credits ........................................... 83

Chart IV.4.1

Retail Loans/GDP and Deviation from Trend ..................................................................................... 87

Chart IV.4.2

Household Financial Assets and Liabilities .......................................................................................... 87

Chart IV.4.3

Average Maturity of General Purpose Loans and ICC ..................................................................... 87

Table IV.4.1

Consumer Loan Maturity Limit .............................................................................................................. 87

Table IV.4.2

ICC Installment Limits ............................................................................................................................. 88

Chart IV.4.4

Change in Credit Card Installment Cap and Installment Expenditures by Sectors ...................... 88

Chart IV.4.5

Credit Card Installment Transactions in Furniture and Decoration Sectors ................................. 89

Chart IV.4.6

Restructured Retail Loans Monitored Under Performing Loans ....................................................... 89

Chart IV.4.7

General Purpose Loan Extensions and Interest Rates ....................................................................... 89

Chart IV.4.8

Loan to Value Ratio for Housing Loans ............................................................................................... 90

Chart IV.4.9

Contributions to Annual Change in Housing Sales ............................................................................ 90

Chart IV.4.10

Housing Loan Extensions and Interest Rates ....................................................................................... 90

Table IV.4.3

General Provision Ratios for Consumer Loans .................................................................................... 90

Chart IV.4.11

Housing and General Purpose Loans .................................................................................................. 91

Chart IV.5.1

Global Liquidity Conditions and Banks’ Non-core Foreign Currency Liabilities ............................. 93

Chart IV.5.2

Global Liquidity Conditions and Banks’ Non-core Foreign Currency Liabilities ............................. 93

Tablo IV.5.1

Summary Statistics .................................................................................................................................. 96

Table IV.5.2

Empirical Results ..................................................................................................................................... 96

Table IV.5.3

Empirical Results ..................................................................................................................................... 97

Table IV.5.4

Empirical Results ..................................................................................................................................... 97

Abbreviations BAT

Banks Association of Turkey

BIS

Bank for International Settlements

BIST

Istanbul Stock Exchange

BKK

Individual Credit Card

BRSA

Banking Regulation and Supervision Agency

CAR

Capital Adequacy Ratio

Central Bank

CBRT

Central Bank of the Republic of Turkey

Credit Card

CDS

Credit Default Swap

CMA

Capital Markets Board of Turkey

CNY

Chinese Yuan

CPI

Consumer Price Index

Discount Factor

DTH

Foreign Exchange Deposit Account

EAD

Exposure at Default

EBITDA

Earnings before Interest Taxes Depreciation and Amortization

ECC

Economic Coordination Committee

ECL

Expected Credit Loss

European Union

FED

Federal Reserve System

FOMC

Federal Open Market Committee

Foreign Exchange

FXNG

FX Net General Position

GDDS

Government Domestic Debt Securities

GDP

Gross Domestic Product

GOĂ&#x153;

Developing Countries

IAS

International Accounting Standards

IASB

International Accounting Standards Board

ICR

Interest Coverage Ratio

IFRS

International Financial Reporting Standards

IFS

International Financial Statistics

IMF

International Monetary Fund

KGF

Credit Guarantee Fund

KKB

Credit Bureau of Turkey

LCR

Liquidity Coverage Ratio

LGD

Loss Given Default

LLW

Late Liquidity Window

Moving Average

MKK

Central Registry Agency

NAFTA

North American Free Trade Agreement

NPL

Non-Performing Loans

NSFR

Net Stable Funding Ratio

Probability of Default

PMC

Pension Monitoring Center

PMI

Purchasing Managers' Indexes

PPS

Private Pension System

Quarter

RHA

Right Hand Axis

RLS

Retail Loan Score

ROA

Return on Assets

ROM

Reserve Option Mechanism

RWA

Risk Weighted Assets

SCT

Special Consumption Tax

SME

Small and Medium-Sized Enterprises

SMEDO

Small and Medium Enterprises Development Organization

TFRS

Turkey Financial Reporting Standards

Turkish Lira

TOBB

Union of Chambers and Commodity Exchanges of Turkey

TOKÄ°

Housing Development Administration of Turkey

TPP

Trans-Pacific Partnership

TURKSTAT Turkish Statistical Institute USA

United States of America

VAT

Value Added Tax

WTO

World Trade Organization