Econophysics Approaches to Large-Scale Business Data and Financial Crisis (eBook)

http://www.smp.dis.titech.ac.jp/en/index.phphttp://www.ier.hit-u.ac.jp/English/faculty/watanabeT.htmlhttp://www.sonycsl.co.jp/en/lab/frl/hideki-takayasu.html

Econophysics Approaches 
 
 
2 
Preface 
6 
Contents 
8 
Contributors 
10 
Part 1 Financial Market Properties 14
Trend Switching Processes in Financial Markets 15
1 Introduction 16
2 Financial Market Data 17
2.1 German Market: DAX Future 17
2.2 US Market: S& P500 Stocks
3 Renormalization Method 19
3.1 Volatility Analysis 21
3.2 Volume Analysis 25
3.3 Inter-trade Time Analysis 28
3.4 Random Shuffling 30
3.5 Universality of Power Law Exponents 33
4 Summary and Conclusions 35
References 37
Nonlinear Memory and Risk Estimation in Financial Records 39
1 Introduction 39
2 Generation of Multifractal Data Series 42
3 Multifractal Analysis 43
4 Return Intervals in the MRC Record 45
5 Return Intervals in Financial Records 47
6 Risk Estimation 49
6.1 Return Interval Approach 49
6.2 Pattern Recognition Technique 51
7 Decision-making Algorithm and the ROC-Analysis 53
8 The Value-at-Risk 56
References 59
Microstructure and Execution Strategies in the Global Spot FX Market 61
1 Introduction 61
2 The Specifics of the EBS Market 63
3 The Microstructure and Dynamics of the Global FX Market 64
3.1 The Structure of the Order Book 64
3.2 Dynamics of the Top of the Order Book 64
3.3 Order Execution Dynamics 66
3.4 Autocorrelations in FX Order Flows and Deal Flows 67
4 Algorithmic Trading of Large Limit Orders 69
5 Simulations of Maker Loss 71
6 Concluding Remarks 73
References 74
Temporal Structure of Volatility Fluctuations 76
1 Introduction 76
2 Data Analyzed 77
3 Scaling and Universality 78
4 Long-Range Correlations 83
5 Summary 86
References 87
Theoretical Base of the PUCK-Model with Application to Foreign Exchange Markets 89
1 Introduction 89
2 The PUCK Model 91
2.1 The Case of Constant b 91
2.2 The Case of Random b(t) 97
3 Statistical Properties of a Real Financial Market 101
4 Summary and Discussions 106
References 108
Part 2 Financial Crisis and Macroeconomics 109
Financial Bubbles, Real Estate Bubbles, Derivative Bubbles, and the Financial and Economic Crisis 110
1 Diagnostics, Proximate and Systemic Origins of the Financial Crisis 111
1.1 Nature of the Financial and Economic Crisis 112
1.2 Standard Explanations for the Financial Crisis 115
1.2.1 Falling Real Estate Values 115
1.2.2 Real-Estate Loans and MBS as a Growing Asset Class Held by Financial Institutions 115
1.2.3 Managers' Greed and Poor Corporate Governance Problem 117
1.2.4 Poor Lending Standards and Deteriorating Regulations and Supervision 118
1.2.5 Did the Fed Cause the Housing Bubble? 119
1.2.6 Bad Quantitative Risk Models in Banks (Basel II) 120
1.2.7 Rating Agency Failures 121
1.2.8 Underestimating Aggregate Risks 122
1.3 The Illusion of the ``Perpetual Money Machine'' 123
2 General Framework for Bubbles and Crashes in Finance 127
2.1 Introduction 127
2.2 Conceptual Framework 129
2.3 Finite-Time Singular Behavior of Bubbles 132
3 A 15-Year History of the 2007--???? Financial and Economic Crisis 135
3.1 First Phase: The ITC ``New Economy'' Bubble (1995--2000) 136
3.2 Second Phase: Slaving of the Fed Monetary Policy to the Stock Market Descent (2000--2003) 138
3.3 Third Phase: Real-Estate Bubbles (2003--2006) 141
3.4 Fourth Phase: MBS, CDOs Bubble (2004--2007) 141
3.5 Fifth Phase: Stock Market Bubble (2004--2007) 145
3.6 Sixth Phase: Commodities and Oil Bubbles (2006--2008) 145
4 Thoughts on Resolution of the Crisis and Its Aftermath 147
4.1 Summary 147
4.2 Trust! Why It Has Been Lost and How to Regain It 149
4.3 Short-term: Melting the Cash Flow Freeze 151
4.4 Long-term: Growth Based on Returning to Fundamentals and Novel Opportunities 152
4.5 The Financial Sphere, Bubbles and Inflation 153
4.6 Recipes for a More Robust and Sustainable World 154
References 155
Global and Local Approaches Describing Critical Phenomena on the Developing and Developed Financial Markets 158
1 Introduction 159
2 Log-Periodicity of Developing Markets: The Case of WIG Index 165
3 Local Fractal Properties of Financial Time Series in the Vicinity of Crashes and Rupture Points 169
4 Exterior Interventions and the Current Situation Seen in Fractal Analysis of Stock Indices 176
5 Conclusions 179
References 180
Root Causes of the Housing Bubble 182
1 Introduction 182
2 A Simple Model of Housing Markets 183
2.1 The Price-to-Rent Ratio 186
3 Determinants of Housing Prices 187
3.1 Price-to-Income Ratio 187
3.2 Elasticity of Housing Supply 188
3.3 The User Cost of Home Ownership 189
3.4 Policies 189
3.5 Home Ownership Rate 190
4 Concluding Remarks 190
References 191
Reconstructing Macroeconomics Based on Statistical Physics 192
1 Introduction 192
2 Is the Statistical Approach Applicable to Economics? 195
3 The Standard Approach Based on the Representative Agent: Some Examples 195
4 The Approach Based on Statistical Physics 198
5 Productivity Dispersion 199
6 Non-self-averaging: A Key Concept for Macroeconomics 202
7 Conclusion 203
References 203
How to Avoid Fragility of Financial Systems: Lessons from the Financial Crisis and St. Petersburg Paradox 205
1 Financial Crisis Viewed as Brittle Fracture 205
2 Diverging Loss in Continuous Trading of Options 207
3 The Saint Petersburg Paradox: Another Solution by Options 209
4 A New Way of Financing Firms: An Alternative Proposal for Financial Firms 211
5 Summary and Discussions 214
References 215
Part 3 General Methods and Social Phenomena 216
Data Centric Science for Information Society 217
1 Change of Society and Scientific Research 217
2 Data Centric Science: A Cyber-Enabled Methodology 218
2.1 Expansion of Research Object and Change in Scientific Methodology Based on ICT 218
2.2 Active Modeling 219
3 Time Series Modeling 220
3.1 State-Space Model and the Kalman Filter 220
3.2 Smoothness Priors Modeling of Time Series 221
4 General State-Space Modeling 223
4.1 Non-Gaussian Filter and Smoother 223
4.2 Monte Carlo Filter and Smoother 224
5 Applications of General State-Space Modeling 225
5.1 Automatic Change Point Detection 225
5.2 Nonlinear Filtering and Smoothing 226
5.3 Data Assimilation 227
5.4 Self-Organizing State-Space Modeling 228
5.5 Semi-Markov Switching Model 229
References 231
Symbolic Shadowing and the Computation of Entropy for Observed Time Series 232
1 Introduction 233
2 Symbolic Dynamics and Shadowing 235
3 Data 241
4 Numerical Results 244
References 250
What Can Be Learned from Inverse Statistics? 252
1 Introduction 252
2 Inverse Statistics and Gain Loss Asymmetry in Financial Indices: A Review 254
2.1 Inverse Statistics 254
2.2 Geometrical Brownian Motion Approximation to Inverse Statistics 255
2.3 Empirical Results 256
2.3.1 Gain--Loss Asymmetry 258
2.4 Possible Causes of the Gain--Loss Asymmetry 259
3 Test of Earlier Results 260
3.1 Emerging Markets 260
3.2 Individual Stocks 260
3.3 Bond Prices 262
4 The Anatomy of GLA 263
4.1 A Contradiction 264
4.2 Extreme Dynamics 264
4.2.1 Fat Tails and Special Events 264
4.2.2 Volatility Scaling of Time 265
5 Regimes 265
5.1 A Simple Non-Mathematical Model 266
5.2 A Mixture Model 267
5.2.1 A Numerical Example 269
5.3 Empirical Findings for the DJIA 270
5.3.1 Three States that Explain the DJIA 270
5.3.2 Comparing Apples and Oranges 270
5.4 Correlations 271
5.4.1 Other Stylized Facts 272
6 Conclusion 273
References 274
Communicability and Communities in Complex Socio-Economic Networks 276
1 Introduction 276
2 The Concept of Communicability 277
3 Communicability and Socio-Economic Communities 280
3.1 Communicability Graph 282
3.2 Overlapping Communities 283
4 Socio-Economic Communities Under External ``Stress'' 284
5 A Social Network Illustration 285
6 Trade Miscellaneous Manufactures of Metal 288
6.1 Analysis of the Communities 289
6.2 Trade Under External Stress 291
7 Conclusions 292
References 293
On World Religion Adherence Distribution Evolution 294
1 Introduction 295
2 Data Bank: Theoretical and Methodological Framework 301
3 Results 302
3.1 Intermediary Comments 302
3.2 Growing Religions 304
3.3 Decaying Religions 305
3.4 Cases with Presently Observed Extremum 307
4 Discussion and Conclusions 312
References 316
Index 318

"Data Centric Science for Information Society (p. 211-212)

Genshiro Kitagawa

Abstract Due to rapid development of information and communication technologies, the methodology of scientific research and the society itself are changing. The present grand challenge is the development of the cyber-enabled methodology for scientific researches to create knowledge based on large scale massive data. To realize this, it is necessary to develop a method of integrating various types of information. Thus the Bayes modeling becomes the key technology. In the latter half of the paper, we focus on time series and present general state-space model and related recursive filtering algorithms. Several examples are presented to show the usefulness of the general state-space model.

1 Change of Society and Scientific Research

By the progress of information and communication technologies (ICT), large-scale massive heterogeneous data have accumulated in various fields of scientific researches and society. As examples, we may consider the microarray data in life science, POS data in marketing, high-frequency data in finance, all-sky CCD image in astronomy, and various data obtained in environmental science and earth science, etc.

These rapid developments changed the society and the research methodologies in science and technology. In the information society, the information became as worthy as the substances and the energy, and the quantity of information was the crucial factor for the success in the society. However, in this twenty-first century, the so-called ubiquitous society is becoming widespread, where everybody can access to huge amount of information anywhere and anytime.

If such ubiquitous society is actually realized, the value of information itself will be depreciated, because everybody can share most information in common. Therefore, the interest in the development of the methods and technologies for information extraction and knowledge creation has grown, because the success and failure in the ubiquitous society depends on whether one can extract essential information from massive data. everybody can share most information in common. Therefore, the interest in the development of the methods and technologies for information extraction and knowledge creation has grown, because the success and failure in the ubiquitous society depends on whether one can extract essential information from massive data.

2 Data Centric Science: A Cyber-Enabled Methodology

2.1 Expansion of Research Object and Change in Scientific Methodology Based on ICT

The scientific research until the nineteenth century has developed basically under Newton–Descartes paradigm based on a mechanic view of the world. In this deductive approach, i.e., in theoretical sciences, mathematics played an important role as the language of science. However, the evolutionism advocated by C. Darwin in mid-nineteenth century concluded that every creature in real world evolves and changes with time. Motivated by such changes of view on the real world, K. Pearson declared in 1891 that everything in the real world can be an object of scientific research, and advocated the grammar of science [15]."