CUTTING-EDGE DEVELOPMENTS IN HIGH-FREQUENCY FINANCIAL
ECONOMETRICS



In recent years, the availability of high-frequency data and
advances in computing have allowed financial practitioners to
design systems that can handle and analyze this information.
Handbook of Modeling High-Frequency Data in Finance
addresses the many theoretical and practical questions raised by
the nature and intrinsic properties of this data.

A one-stop compilation of empirical and analytical research,
this handbook explores data sampled with high-frequency finance in
financial engineering, statistics, and the modern financial
business arena. Every chapter uses real-world examples to present
new, original, and relevant topics that relate to newly evolving
discoveries in high-frequency finance, such as:

* Designing new methodology to discover elasticity and plasticity
of price evolution

* Constructing microstructure simulation models

* Calculation of option prices in the presence of jumps and
transaction costs

* Using boosting for financial analysis and trading

The handbook motivates practitioners to apply high-frequency
finance to real-world situations by including exclusive topics such
as risk measurement and management, UHF data, microstructure,
dynamic multi-period optimization, mortgage data models, hybrid
Monte Carlo, retirement, trading systems and forecasting, pricing,
and boosting. The diverse topics and viewpoints presented in each
chapter ensure that readers are supplied with a wide treatment of
practical methods.

Handbook of Modeling High-Frequency Data in Finance is an
essential reference for academics and practitioners in finance,
business, and econometrics who work with high-frequency data in
their everyday work. It also serves as a supplement for risk
management and high-frequency finance courses at the
upper-undergraduate and graduate levels.

Frederi G. Viens, PhD, is Director and Coordinator of the
Computational Finance Program at Purdue University, where he also
serves as Professor of Statistics and Mathematics. He has published
extensively in the areas of mathematical finance, probability
theory, and stochastic processes. Dr. Viens is co-organizer of the
annual Conference on Modeling High-Frequency Data in Finance.

Maria C. Mariani, PhD, is Pro-fessor and Chair in the
Department of Mathematical Sciences at The University of Texas at
El Paso. She currently focuses her research on mathematical
finance, applied mathematics, and numerical methods. Dr. Mariani is
co-organizer of the annual Conference on Modeling High-Frequency
Data in Finance.

Ionut Florescu, PhD, is Assistant Professor of
Mathematics at Stevens Institute of Technology. He has published in
research areas including stochastic volatility, stochastic partial
differential equations, Monte Carlo methods, and numerical methods
for stochastic processes. Dr. Florescu is lead organizer of the
annual Conference on Modeling High-Frequency Data in Finance.

In recent years, the availability of high-frequency data and advances in computing have allowed financial practitioners to design systems that can handle and analyze this information. Handbook of Modeling High-Frequency Data in Finance addresses the many theoretical and practical questions raised by the nature and intrinsic properties of this data.

A one-stop compilation of empirical and analytical research, this handbook explores data sampled with high-frequency finance in financial engineering, statistics, and the modern financial business arena. Every chapter uses real-world examples to present new, original, and relevant topics that relate to newly evolving discoveries in high-frequency finance, such as:

• Designing new methodology to discover elasticity and plasticity of price evolution

• Constructing microstructure simulation models

• Calculation of option prices in the presence of jumps and transaction costs

• Using boosting for financial analysis and trading

The handbook motivates practitioners to apply high-frequency finance to real-world situations by including exclusive topics such as risk measurement and management, UHF data, microstructure, dynamic multi-period optimization, mortgage data models, hybrid Monte Carlo, retirement, trading systems and forecasting, pricing, and boosting. The diverse topics and viewpoints presented in each chapter ensure that readers are supplied with a wide treatment of practical methods.

Handbook of Modeling High-Frequency Data in Finance is an essential reference for academics and practitioners in finance, business, and econometrics who work with high-frequency data in their everyday work. It also serves as a supplement for risk management and high-frequency finance courses at the upper-undergraduate and graduate levels.

Preface xi

Contributors xiii

Part One Analysis of Empirical Data 1

1 Estimation of NIG and VG Models for High Frequency Financial Data 3
José E. Figueroa-López Steven R. Lancette Kiseop Lee and Yanhui mi

1.1 Introduction 3

1.2 The Statistical Models 6

1.3 Parametric Estimation Methods 9

1.4 Finite-Sample Performance via Simulations 14

1.5 Empirical Results 18

1.6 Conclusion 22

References 24

2 A Study of Persistence of Price Movement using High Frequency Financial Data 27
Dragos Bozdog Ionu Florescu Khaldoun Khashanah and Jim Wang

2.1 Introduction 27

2.2 Methodology 29

2.3 Results 35

2.4 Rare Events Distribution 41

2.5 Conclusions 44

References 45

3 Using Boosting for Financial Analysis and Trading 47
Germán Creamer

3.1 Introduction 47

3.2 Methods 48

3.3 Performance Evaluation 53

3.4 Earnings Prediction and Algorithmic Trading 60

3.5 Final Comments and Conclusions 66

References 69

4 Impact of Correlation Fluctuations on Securitized structures 75
Eric Hillebrand Ambar N. Sengupta and Junyue Xu

4.1 Introduction 75

4.2 Description of the Products and Models 77

4.3 Impact of Dynamics of Default Correlation on Low-Frequency Tranches 79

4.4 Impact of Dynamics of Default Correlation on High-Frequency Tranches 87

4.5 Conclusion 92

References 94

5 Construction of Volatility Indices Using A Multinomial Tree Approximation Method 97
Dragos Bozdog Ionu Florescu Khaldoun Khashanah and Hongwei Qiu

5.1 Introduction 97

5.2 New Methodology 99

5.3 Results and Discussions 101

5.4 Summary and Conclusion 110

References 115

Part Two Long Range Dependence Models 117

6 Long Correlations Applied to the Study of Memory Effects in High Frequency (TICK) Data the Dow Jones Index and International Indices 119
Ernest Barany and Maria Pia Beccar Varela

6.1 Introduction 119

6.2 Methods Used for Data Analysis 122

6.3 Data 128

6.4 Results and Discussions 132

6.5 Conclusion 150

References 160

7 Risk Forecasting with GARCH Skewed t Distributions and Multiple Timescales 163
Alec N. Kercheval and Yang Liu

7.1 Introduction 163

7.2 The Skewed t Distributions 165

7.3 Risk Forecasts on a Fixed Timescale 176

7.4 Multiple Timescale Forecasts 185

7.5 Backtesting 188

7.6 Further Analysis: Long-Term GARCH and Comparisons usi…