Realistic Simulation of Financial Markets (eBook)

Foreword 6
Preface 8
Acknowledgements 14
Contents 16
Part I U-Mart System: The First Test Bed of the Third Mode of Science 17
1 A Guided Tour of the Backside of Agent-Based Simulation 18
1.1 Introduction 18
1.2 General Crisis of Economics: State of Economics During and Before the First Half of the 1970s 21
1.2.1 Capital Theory Controversies 22
1.2.2 Marginal Cost Controversy 26
1.2.3 ``Empty Boxes'' Controversy and Sraffa's Analysis on Laws of Returns 30
1.3 Possibilities and Limits of General Equilibrium: State of Economics After the 1970s 33
1.3.1 Assumptions of Arrow and Debreu's Formulation 34
1.3.2 Problems of Interpretations of the Arrow-Debreu Theory 38
1.3.3 Shapes of Excess Demand Functions 39
1.3.4 GET and Increasing Returns to Scale 41
1.3.5 Computable General Equilibrium 45
1.3.6 Dynamic Stochastic General Equilibrium Models 46
1.3.7 Why did the Mainstream Research Program Fail? 47
1.3.8 What Happened During this Century and a Half? 49
1.4 Tasks and Possibilities of ABS 50
1.4.1 New Bag for a New Wine: ABS as a New Tool for Economics 50
1.4.2 The Third Paradigm in Scientific Research 53
1.4.3 Complexity and Tractability 55
1.4.4 Features of Human Behavior 57
1.4.5 Evolutionary Economics and Micro-Macro Loops 59
1.5 Conclusions 61
References 62
2 Research on ABS and Artificial Market 66
2.1 Social Simulation 66
2.1.1 Methods for Social Simulation 66
2.1.1.1 System Dynamics: Models Using Macroscopic Variables 66
2.1.1.2 Models Using Microscopic Variables 67
2.1.2 Research with Simulation 68
2.2 Agent-Based Simulation 68
2.2.1 Structure of the Agent-Based Models 68
2.2.1.1 Implementation Issues of Agent 69
2.2.1.2 Interaction Structure of Agents 69
2.2.2 Simulator Granularity 70
2.3 Hybridization with Gaming 71
2.4 Characteristics of U-Mart as an Agent-Based Simulation Model 72
References 73
3 Building Artificial Markets for Evaluating Market Institutions and Trading Strategies 74
3.1 The Fidelity of Models: From KISS Principle to High-Fidelity Models 74
3.2 Requirements for Artificial Market Simulators for Evaluating Market Institutions and Trading Strategies 76
3.3 Itayose U-Mart System (U-Mart System Ver.2) 76
3.3.1 System Configuration 77
3.3.1.1 Time Management in the Itayose U-Mart System 77
3.3.1.2 The Transaction Method in the Itayose U-Mart System 78
3.3.1.3 GUI Tools of the Itayose U-Mart System 79
3.3.2 Implementation of Itayose Market Server 80
3.3.3 How to Develop Trading Agents for Itayose U-Mart System 81
3.3.4 Features and Problems of the Itayose U-Mart System 85
3.4 Zaraba-Based U-Mart System (U-Mart System Ver.4) 86
3.4.1 System Configuration 87
3.4.1.1 Time Management in the Zaraba-Based U-Mart System 87
3.4.1.2 The Transaction Method in the Zaraba-Based U-Mart System 88
3.4.1.3 GUI Tools of the Zaraba-Based U-Mart System 91
3.4.2 Implementation of Zaraba Market Server 91
3.4.2.1 The Exchange Module 91
3.4.2.2 The Security Company Module 93
3.4.3 How to Develop Trading Agents for the Zaraba-Based U-Mart System 93
3.4.4 Features 97
3.5 An Example of Numerical Experiments: An Effect of a Futures Market on Its Spot Market 97
3.5.1 Experimental Settings 98
3.5.2 Results 99
References 100
4 A Perspective on the Future of the Smallest Big Project in the World 101
4.1 Characteristics of a Big Project and U-Mart 101
4.2 Agent-Based Modeling Toward New Social System Sciences 102
4.2.1 Requirements on Agent-Based Simulation Models 103
4.2.2 Toward a New Research Scheme for Agent-Based Simulation in Social and Economic Complex Systems 105
4.3 Concluding Remarks 106
References 107
Part II Applications of Artificial Markets 108
5 Evolution of Day Trade Agent Strategy by Means of Genetic Programming with Machine Learning 109
5.1 Introduction 109
5.2 Day Trading 110
5.2.1 Advantage of Day Trading 110
5.2.2 Order Book Information 110
5.2.2.1 Spread 111
5.2.2.2 Thickness 111
5.2.3 Technical Analysis 112
5.3 Day Trade Agent Framework 112
5.3.1 Market Replay System 112
5.3.1.1 Data-Collecting Part 112
5.3.1.2 Data-Providing Part 113
5.3.2 Agent Learning Part 113
5.3.3 Trading Agent 113
5.4 Genetic Programming and Support Vector Machine 114
5.4.1 Genetic Programming 114
5.4.2 Support Vector Machine 114
5.5 Evolution of the Day Trading Strategy by Genetic Programming 115
5.5.1 Individual Expression of the Day Trade Strategy 115
5.5.2 Action of the Trade Agent 117
5.5.3 Fitness Measure 117
5.5.4 Learning Term 118
5.6 Propose Method 119
5.6.1 Prediction of Fluctuation of Stock Prices with SVMs 119
5.6.2 Evolution of Trading Strategy by GP 120
5.6.2.1 The Method to Apply the Individual of GP 120
5.6.2.2 Nodes of GP Individual 120
5.7 Computer Experiments 121
5.7.1 Experiment Condition 121
5.7.2 The Prediction of the Fluctuation of Stock Prices with SVMs 122
5.7.3 The Strategy in GP 123
5.8 Conclusion 126
References 126
6 How to Estimate Market Maker Models in an Artificial Market 128
6.1 Introduction 128
6.2 Thin Market 130
6.3 Market Maker 133
6.3.1 Three Models of Simple Market Maker 133
Notations listed blow are used to represent models 134
6.3.2 Market Maker Model 1 (MM1, Simple Spread Type) 134
6.3.3 Market Maker Model 2 (MM2, Linear Type) 135
6.3.4 Market Maker Model 3 (MM3, Polynomial Type) 136
6.4 Efficiency and Feasibility of Market Maker Models 137
6.4.1 Contract Rate of a Market with Random Agents 137
6.4.2 Experimental Environments Given by the U-Mart Artificial Futures Market 138
6.4.3 Results of Experiment 1 (In Geometric Brownian Motion) 139
6.4.4 Results of Experiment 2 (GARCH Process) 141
6.4.5 Human Agent (U-Mart 2005) 144
6.5 Conclusion 145
References 146
7 The Effect of Resilience in Optimal Execution with Artificial-Market Approach 147
7.1 Introduction 147
7.1.1 Background 147
7.1.2 Objective 148
7.1.3 Meaning of Using U-Mart in This Preset Research 149
7.2 Optimal Investment Strategies in Financial Markets 150
7.2.1 What Is an Optimal Investment Strategy? 150
7.2.2 Classification of the Preceding Studies 151
(1) Goals 152
(2) Period 152
(3) Decision-Making Timing 152
(4) Market-Related Hypotheses 153
7.2.3 Introduction of Optimal Investment Strategies of Risk-Neutral Heavy Traders 153
7.2.3.1 Bertsimas and Lo Model 153
7.2.3.2 Obizhaeva and Wang Model 154
7.2.3.3 Comparison of Two Strategies Based on Numerical Samples 157
7.2.4 Is It Appropriate to Give Resilience Exogenously? 161
7.3 Resilience in the Optimal Investment Strategy 162
7.3.1 What Is Resilience? 162
7.3.2 Meaning of Considering Resilience in the Optimal Investment Strategy 163
7.3.3 Proposal of Resilience Models 163
7.3.3.1 Resilience Brought About by Trader Investment Behaviors of Traders 163
7.3.3.2 Zero Intelligence Model 164
7.3.3.3 Full Intelligence Model 165
7.3.3.4 Low Intelligence Model 165
7.3.4 How Should Resilience Be Quantified? 166
7.3.4.1 Resilience Movements 166
7.3.4.2 Quantification of Resilience Conducted by Christalla 166
7.3.4.3 Evaluation of Three Models Based on Resilience 167
7.4 Analysis of Optimal Investment Strategies Based on the U-Mart System 169
7.4.1 Simulation Overview 169
7.4.2 Simulation Results 170
7.4.3 Consideration 176
7.5 Conclusion 177
7.5.1 Future Issues 178
References 178
8 Observation of Trading Process, Exchange, and Market 180
8.1 Introduction 180
8.2 Trading Processes in the Stock Market 182
8.2.1 The Call Auction and the Continuous Double Auction 182
8.2.2 Observation of Trading Processes by Artificial Market Experiments 183
8.2.2.1 Experiments in 2013 185
8.2.2.2 Experiments in 2014 186
8.3 Contract and the Principal of Exchange 190
8.3.1 The Importance of Contract 190
8.3.2 The Principle of Exchange 191
8.3.3 Exchange with Money 192
8.3.4 Quotes and Prices 193
8.4 Arbitrage Trading and Maket 195
8.4.1 Arbitrage Trading 195
8.4.2 Different Decisions and Market 197
8.5 Conclusions 198
8.6 Mathematical Notices 200
References 201
Index 203