Intelligent Trading Systems: Applying Artificial Intelligence to Financial Markets

by Ondrej Martinsky

Overview

This book deals with the issue of problematic market price prediction in the context of crowd behaviour affected by the psychology of the masses. It highlights the contrast between a phenomenon of mass psychology and the efficient market hypothesis, which is essentially based on a common economic theory. The basic assumption is that if there is a model of interaction between masses and agents participating in markets, then there also exist means for prediction of the whole market's behaviour, though nevertheless the behaviour of every single agent is not predictable.
From a practical point of view, this book describes technical analysis methods used to predict price movements, and discusses a soft computing approach used in a composition of automated trading systems. This book brings alternative, soft computing computational models to trading strategies and innovatively combines two different areas of science - artificial intelligence and technical analysis. One of the main benefits of this book is a demonstration that the soft computing approach in a combination with the "soft" social sciences accounts more reliable results than the conventional mathematical models.
This book is for anyone interested in trading, financial markets and security exchanges, as well as for those who have theoretical or practical knowledge from the fields of artificial intelligence and soft computing, and want to know how these topics can be applied in financial markets.

Product Details

ISBN-13:	9781906659530
Publisher:	Harriman House Publishing
Publication date:	01/01/2010
Pages:	212
Product dimensions:	6.40(w) x 8.90(h) x 0.70(d)

About the Author

Ondrej Martinsky specialises in mathematics and computer science, intelligent and decision support systems. He has been working on several research projects within these areas during his stay at Brno University of Technology. He currently works as an independent contractor for a major investment bank in London and publishes his blog www.quantandfinancial.com. Ondrej is also the author of several scientific papers which have been presented at international conferences.
Ondrej also studies economics, financial markets and security exchanges, and in particular technical analysis of market prices in combination with the short-term trading of financial derivatives. Alongside conventional trading he is also interested in the scientific approach, wondering not only how markets move, but also why they move as they do.
During the last few years, he has been investigating how advanced computational methods can be applied in financial markets. In his book, "Intelligent Trading Systems", he provides the results of this research and a unique interconnection of knowledge from the fields of computer science and financial markets.

Table of Contents

Introduction 1

1 Reality, the intersection of multiple theories 7

1.1 Efficient market hypothesis 8

1.2 The theory of chaos 10

1.3 Behavioral market theory 14

2 The dynamics of crowd behavior 19

2.1 Methodologies for the study of markets 19

2.2 The system theory point of view 21

2.2.1 The exchange of energy and information 22

2.2.2 The crowd's life cycle 24

2.2.3 Unexpected events and shocks 27

2.2.4 Generalized turnover patterns 29

2.2.5 Generalized pro-trend patterns 32

2.3 The wave principle 35

2.3.1 The hierarchical organization of Elliott waves 37

2.3.2 The direction of waves 38

2.3.3 The mode of waves 39

2.3.4 The hierarchy of complete cycles 42

2.3.5 Variations of motive waves 44

2.3.6tVariations of corrective waves 48

2.3.7 The principle of alternation 52

2.4 Fibonacci mathematics in financial markets 52

2.4.1 The golden ratio 53

2.4.2 The golden rectangle and golden spiral 54

2.4.3 The application of Fibonacci numbers in financial markets 55

3 Security exchanges at a glance 61

3.1 Financial markets 61

3.2 Security exchanges 63

3.2.1 Entities participating in markets 63

3.2.2 Order-driven and quote-driven markets 68

3.2.3 World's largest and most long-standing security exchanges 70

3.2.4 Types of orders 72

3.2.5 Pit trading versus electronic trading 76

3.3 Exchange clearing systems 78

4 Basic tenets of automated trading 85

4.1 Indicators and oscillators 86

4.1.1 Moving averages 86

4.1.2 Average directional index 90

4.1.3 Average true range 92

4.1.4 Relative strength index 93

4.1.5 Bollinger bands 95

4.2 Money management 96

4.3 Statistics 100

4.4 The sensitivity to changes of parameters 104

5 Simulation and backtesting of trading strategies 107

5.1 The value of simulation in trading 107

5.2 Human factor in the trading chain 108

5.3 Modeling of intra-bar price movements 109

5.4 Modeling of order execution 111

5.5 Modeling of time and price skews 113

5.6 Discrete Event System Specification 115

5.6.1 DEVS formalism 116

5.6.2 Simulators and coordinators for DEVS 117

5.7 Simulation of the trading environment 122

5.7.1 The data provider component 125

5.7.2 The delay component 126

5.7.3 The order execution component 127

5.7.4 The ATS component 129

5.7.5 The parallel run of multiple trades 133

5.8 Embedding trading strategies into the simulation 135

5.8.1 Simulation case study 136

6 Optimization of trading strategies 141

6.1 Parametric trading strategies 141

6.1.1 Choosing an appropriate fitness function 143

6.1.2 Parametric surface 146

6.2 Exhaustive search 149

6.3 Genetic algorithms 150

6.3.1 Inspiration from nature 151

6.3.2 Computational model of genetic evolution 151

6.3.3 Optimization case study 157

7 Fuzzy approach to trading strategies 161

7.1 Concept of uncertainty and the basics of fuzzy logic theory 162

7.1.1 Linguistic variables and fuzzy sets 163

7.2 Fuzzy logic and fuzzy inference 166

7.2.1 Fuzzification 168

7.2.2 Inference engine and evaluation of rules 169

7.2.3 Defuzzification 176

7.3 Fuzzy-based trading strategies 177

7.3.1 Triple Screen Trading System 177

7.3.2 Fuzzy approach to the Triple Screen Trading System 180

7.4 Analysis of sensitivity and robustness 187

7.4.1 Sensitivity analysis of the whole market system 187

7.4.2 Sensitivity analysis of the signaling system 188

7.5 Case study 190

Summary 193

Bibliography and further reading 195

Notations, functions and mathematical symbols 199