FOREX Perfection
In Manual, Automated And Predictive Trading
MOSTAFA AFSHARI
First Edition
CopyrightŠ2018 by Mostafa Afshari All rights reserved. No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author, except in the case of brief quotations embodied in critical reviews and certain other noncommercial uses permitted by copyright law. For permission requests or sending feedbacks to be published in the next editions please contact the author at mostafaafshari@hushmail.com
DISCLAIMER The purpose of this book is mainly informational. The user of this book is the sole responsible for any possible loss that may incur by using the information herein. The author totally declaims such losses.
Table of Contents Introduction Manual Trading Automated Trading Predictive Trading Chapter One: Trading Signals 1.1 Highly Probable Chart Patterns 1.2 Price Rejection 1.3 Deviation From Mean 1.4 Volume Up And Down Spikes Chapter Two: Essentials of Manual Trading 2.1 Signal Strength (Accuracy) 2.2 Signal Profitability 2.3 Signal Manageability Chapter Three: Scheduling A Manual Trade Chapter Four: Automated Trading Platforms Chapter Five: Introduction to Automated Trading Strategies Chapter Six: Developing an Automated Strategy 6.1 Choosing Base Signal 6.2 Choosing Entry And Exit Methods 6.3 Pre-Processing (Smoothing or Filtering) of Base Signal(s) 6.4 Backtesting, Optimization, Performance Evaluation And Unseen Data (Walk Forward) Testing 6.5 Strategy Diagnosis
1 1 2 3 6 6 16 18 22 25 25 29 31 33 39 46 50 50 59 60 60 65
Table of Contents Chapter Seven: Demo Trading Hiring A VPS Chapter Eight: Real Automated Trading Chapter Nine: Modeling of Forex Time Series 9.1 Types of Forex Data 9.2 Model Execution and Optimization (BCMEM) Chapter Ten: Prediction Models 10.1 Forex Prediction Using Gaussian Process 10.2 Forex Prediction Using Analytical Functions 10.3 Holy Grail of Forex Prediction Chapter Eleven: 15 Minutes A Day Trading Chapter Twelve: Final Advice
73 74 77 77 80 82 82 90 93 95 96
1
Introduction At first look, doing Forex seems too easy and straightforward but in reality many difficulties emerge. In my opinion, the reason is that Forex is a dynamic system and managing such systems is quite tricky. However, in this book I have tried to provide some qualified, statistically sound solutions to beat this hard to manage system. My solutions have been organized in three categories: manual, automated and predictive trading. Traders, according to their habits and trading style can pick one, two or a mix of the methods to finally do Forex as flawless as possible. Manual Trading In this chapter I have discussed manual trading around three essential concepts of accuracy, profitability and manageability. Using these concepts I have shown how filter out bad trades and reach handpicked super highly probable trades to execute. The following picture shows an example of this process of high performance trading.
2
Automated Trading Thanks to the modern technologies, automated and algorithmic trading is getting more popular day by day. In this part of the book, using a super-fast, easy, flexible and advanced free software you will learn how develop viable automated strategies in just a few days. The software language is super easy to be learned and you can make use of it in a few days. Moreover, some new thrilling features are added to this software in each release that makes it even more powerful. I highly recommend you don't miss this stunning software and its capabilities.
3
The following picture shows backtest results of a sample strategy that has been developed by this software:
Predictive Trading One of the biggest difficulties in trading is lack of confidence and certainty to enter a trade. Because you cannot thoroughly analyze historical charts, you are in constant hesitation to enter your expected trades. Imagine you can predict next day prices (with around 80% accuracy) using mathematical methods and decide to trade according to these predictions. Having this scientific backup gives you big courage and conf-
4
-idence to trade and makes your trade highly creditable. Believe me or not, this almost impossible job has been done in this book and now is ready for you to be used. If your computer skills allow you to just install Python or R (as easy as any other one click installation software), you are there to go. Besides ordinary traders, the methods I have explained in this chapter may be useful for finance researchers, institutional traders, algorithmic trading developers and any person who want to learn one or two lessons about modeling financial time series. As a sample, the following picture shows prediction of EUR/USD price for 16th November 2017 using a Python package!
5
PART I. MANUAL TRADING
20
38
PART II. AUTOMATED TRADING
64
Return, Profit Factor, Sharpe Ratio, Kelly Criterion, R2 Coefficient and Ulcer Index. These six vital performance metrics plus Max Drawdown and Percent Winning rate are the only performance criteria you need to judge about a strategy. In general, max drawdown less than 10%, profit factor greater than 1.50, Sharpe and Kelly greater than 2.00, percent winning better than 65%, R2 better than 70%, Ulcer Index less than 10% and indeed annual return greater than 100% are noticeable. These performance values are the minimum threshold for a prospective strategy. However, in reality for a strategy to be profitable, we need to heavily optimize it and the performance metrics must be as great as possible. A performance result page:
71
75
PART III. PREDICTIVE TRADING
87
may be related to your Anaconda installation. Uninstall Anaconda Python package and try installation steps again. If after all these steps you still have problems reaching Python Spyder IDE, you need to consult Google to resolve the problem or install another Python distribution which includes Spyder from here: https:// pythonhosted.org/spyder/installation.html On the following picture you can find a screenshot of Spyder console:
R Installation In comparison to Python, installing and using R is too way easier. Just download R and its famous IDE from here https://www.r-project.org
89
Besides Python, MATLAB users can find MATLAB codes to predict Forex prices using MATLAB Gaussian Process tools (in Statistics & Machine Learning Toolbox). To have GP you need to install MATLAB 2017 or later releases. MATLAB provides you countless options and capabilities to model using GP. You can find my MATLAB sample codes in the downloaded data and codes (see section 10.1 or chapter four for download links). The following table shows typical price predictions for 20 days (EUR/USD and AUD/ USD) using Python Gaussian Process packages. The stared values are those that predicted and actual have big difference and are not usable as accurate predictions in real trading.
After reading a sample of “FOREX Perfection In Manual, Automated and Predictive Trading� you may have been convinced that this book have something valuable to learn. To recap what you will get from this book: 1. How hunt 90% probable signals and place top hand-picked high performance trades. 2. How fully automate your trading using a free, smart, advanced piece of software and start making money on autopilot. 3. How predict (yes you read correctly!) daily prices (with 80% accuracy) using a scientific method and trade according to this prediction. Get all these unheard methods of FOREX trading in this book here now: http://bit.ly/2Q9Z2Ar