Contents
1 Introduction 1
1.1 The London Stock Exchange and the LSE data . . . . 2
1.1.1 Trading day . . . . . . . . . . . . . . . . . . . . 2
1.1.2 Member firms . . . . . . . . . . . . . . . . . . . 4
1.1.3 The LSE dataset, preparation and cleaning . . 5
1.2 Summary of chapters . . . . . . . . . . . . . . . . . . . 6
2 The power of patience: A behavioral regularity in
limit order placement 9
2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 9
2.2 Description of the London Stock Exchange data . . . . 11
2.3 Properties of relative limit order prices . . . . . . . . . 12
2.3.1 Unconditional distribution . . . . . . . . . . . . 12
2.3.2 Time series properties . . . . . . . . . . . . . . 15
2.4 Volatility clustering . . . . . . . . . . . . . . . . . . . . 17
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . 19
3 The predictive power of zero intelligence in financial
markets 21
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1.1 Continuous double auction . . . . . . . . . . . 22
3.1.2 Review of the model . . . . . . . . . . . . . . . 24
3.1.3 Predictions of the model . . . . . . . . . . . . . 25
3.2 Testing the scaling laws . . . . . . . . . . . . . . . . . 26
3.2.1 Data . . . . . . . . . . . . . . . . . . . . . . . . 26
3.2.2 Testing procedure . . . . . . . . . . . . . . . . 27
3.2.3 Spread . . . . . . . . . . . . . . . . . . . . . . . 28
3.2.4 Price diffusion rate . . . . . . . . . . . . . . . . 28
3.3 Average market impact . . . . . . . . . . . . . . . . . 29
3.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 34
3.5 Supplementary Material . . . . . . . . . . . . . . . . . 37
3.5.1 Literature review . . . . . . . . . . . . . . . . . 37
3.5.2 Dimensional analysis . . . . . . . . . . . . . . . 40
3.5.3 The London Stock Exchange (LSE) data set . . 41
3.5.4 Opening auction, real order types, time . . . . 43
3.5.5 Measurement of model parameters . . . . . . . 44
3.5.6 Estimating the errors for the regressions . . . . 48
3.5.7 Market impact . . . . . . . . . . . . . . . . . . 52
3.5.8 Extending the model . . . . . . . . . . . . . . . 56
4 Correlation and clustering in the trading of the members
of the LSE 59
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 59
4.1.1 The LSE dataset . . . . . . . . . . . . . . . . . 60
4.1.2 Measuring correlations between strategies . . . 60
4.2 Significance and structure in the correlation matrices . 64
4.2.1 Density of the correlation matrix eigenvalue
distribution . . . . . . . . . . . . . . . . . . . . 64
4.2.2 Bootstrapping the largest eigenvalues . . . . . 67
4.2.3 Clustering of trading behaviour . . . . . . . . . 68
4.2.4 Time persistence of correlations . . . . . . . . . 71
4.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 75
5 Market imbalances and stock returns: heterogeneity
of order sizes at the LSE 77
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . 77
5.2 Distribution of order sizes . . . . . . . . . . . . . . . . 80
5.3 Order size heterogen. and stock returns . . . . . . . . 82
5.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 94
6 Conclusions 97