Gaining Predictive Edge in the Market To order this title, and for more information, click here
By Paul McNelis, Robert Bendheim Professor of International Economic and Financial Policy at Fordham University Graduate School of Business. Professor of Economics at Georgetown University until 2004.
Description This book explores the intuitive appeal of neural networks and the genetic algorithm in finance. It demonstrates how neural networks used
in combination with evolutionary computation outperform classical econometric methods for accuracy in forecasting, classification and
dimensionality reduction.
McNelis utilizes a variety of examples, from forecasting automobile production and corporate bond spread,
to inflation and deflation processes in Hong Kong and Japan, to credit card default in Germany to bank failures in Texas, to cap-floor
volatilities in New York and Hong Kong.
Audience
Upper division undergraduates and MBA students, as well as the rapidly growing number of financial engineering programs, whose curricula
emphasize quantitative applications in financial economics and markets
Contents Preface
1 Introduction
1.1 Forecasting, Classification and Dimensionality Reduction
1.2 Synergies
1.3 The Interface Problems
1.4 Plan
of the Book
Econometric Foundations
2 What Are Neural Networks
2.1 Linear Regression Model
2.2 GARCH Nonlinear Models
2.2.1 Polynomial
Approximation
2.2.2 Orthogonal Polynomials
2.3 Model Typology
2.4 What Is A Neural Network
2.4.1 Feedforward Networks
2.4.2 Squasher
Functions
2.4.3 Radial Basis Functions
2.4.4 Ridgelet Networks
2.4.5 Jump Connections
2.4.6 Multilayered Feedforward Networks
2.4.7 Recurrent
Networks
2.4.8 Networks with Multiple Outputs
2.5 Neural Network Smooth-Transition Regime-Switching Models
2.5.1 Smooth Transition Regime
Switching Models
2.5.2 Neural Network Extensions
2.6 Nonlinear Principal Components: \ Intrinsic Dimensionality
2.6.1 Linear Principal
Components
2.6.2 Nonlinear Principal Components
2.6.3 Application to Asset Pricing
2.7 Neural Networks and Discrete Choice
2.7.1 Discriminant
Analysis
2.7.2 Logit Regression
2.7.3 Probit Regression
2.7.4 Weibull Regression
2.7.5 Neural Network Models for Discrete Choice
2.7.6
Models with Multinomial Ordered Choice
Criticism and Data Mining
2.9 Conclusion
2.9.1 Matlab Program Notes
2.9.2 Suggested Exercises
3 Estimation of a Network with Evolutionary Computation
3.1 Data Preprocessing
3.1.1 Stationarity: Dickey-Fuller Test
3.1.2 Seasonal
Adjustment: Correction for Calendar Effects
3.1.3 Scaling of Data
3.2 The Nonlinear Estimation Problem
3.2.1 Local Gradient-Based Search:
\ The Quasi- Backpropagation 46
Simulated Annealing 48
3.2.3 Evolutionary Stochastic Search: The Genetic Algorithm
Population creation
Selection
Crossover
Mutation
Election tournament
Elitism
Convergence
3.2.4 Evolutionary Genetic Algorithms
3.2.5 Hybridization: Coupling
Gradient- and Genetic Search Methods
3.3 Repeated Estimation and Thick Models
3.4 Matlab Examples: Numerical Performance 53
3.4.1 Numerical
Optimization
3.4.2 Approximation with Networks 54
3.5 Conclusion
3.5.1 Matlab Program Notes
3.5.2 Suggested Exercises
4 Evaluation
of Network Estimation
4.1 In-Sample Criteria
4.1.1 Goodness of Fit Measure
4.1.2 Hannan-Quinn Information Criterion
4.1.3 Serial Independence
and Homoskedasticity: and McLeod-Li Tests
4.1.4 Symmetry
Normality
4.1.6 Neural Network Test for Neglected Nonlinearity: Lee-White-Granger
Test
4.1.7 Brock-Deckert-Scheinkman Test for Nonlinear Patterns
4.1.8 Summary of in-sample criteria
4.1.9 Matlab Example
4.2 Out-of-Sample
Criteria
4.2.1 Recursive Methodology
4.2.2 Root Mean Squared Error Statistic
4.2.3 Diebold-Mariano Test for Out of Sample Errors
4.2.4
Harvey, Leybourne, and Newbold "Size Correction" of Diebold-Mariano Test
4.2.5 Out-of-Sample Comparison with Nested Models
4.2.6 Success
Ratio for Sign Predictions: Directional Accuracy
4.2.7 Predictive Stochastic\ Complexity
subsection \numberline 4.2.8 Cross-Validation
and the Method 69
How Large for Predictive Accuracy
4.3 Interpretive Criteria and Significance of Results
4.3.1 Analytic Derivatives
4.3.2 Finite Differences
4.3.3 Does It Matter
4.3.4 Matlab Example: Analytic and Finite Differences
4.3.5 Bootstrapping for Assessing
Significance
4.4 Implementation Strategy
4.5 Conclusion
4.5.1 Matlab Program Notes
4.5.2 Suggested Exercises
1em Applications and Examples
5 Estimation and Forecasting with Artificial Data
5.1 Introduction
5.2 Stochastic Chaos Model
5.2.1 In-Sample Performance
5.2.2 Out-of-Sample
Performance
5.3 Stochastic Volatility/Jump Diffusion Model
5.3.1 In-Sample Performance
5.3.2 Out-of-Sample Performance
5.4 The Markov
Regime Switching Model
5.4.1 In-Sample Performance
5.4.2 Out-of-Sample Performance
5.5 VRS Model
5.5.1 In-Sample Performance
5.6 Distorted
Long Memory Model
5.6.1 In-Sample Performance
5.6.2 Out-of-Sample Performance
5.7 BSOP Model: Implied Volatility Forecasting
5.7.1 In-Sample
Performance
5.7.2 Out-of-Sample Performance
5.8 Conclusion
5.8.1 Matlab Program Notes
5.8.2 Suggested Exercises
6 Times Series: Examples
from Industry and Finance
6.1 Forecasting Production in the Automotive Industry
6.1.1 The Data
6.1.2 Models of Quantity Adjustment
6.1.3
In-Sample Performance
6.1.4 Out-of-Sample Performance
6.1.5 Interpretation of Results
6.2 Corporate Bonds: Which Spreads? 110
6.2.1
The Data
6.2.2 A Model for the Adjustment of Spreads
In-Sample Performance
6.2.4 Out-of-Sample Performance
6.2.5 Interpretation of Results
6.3 Conclusion
6.3.1 Matlab Program Notes
6.3.2 Suggested Exercises
7 Inflation and Deflation: Hong Kong and Japan
7.1 Hong Kong
7.1.1
The Data
7.1.2 Model Specification
7.1.3 In-Sample Performance
7.1.4 Out-of-Sample Performance
7.1.5 Interpretation of Results
7.2 Japan
7.2.1 The Data
7.2.2 Model Specification
7.2.3 In-Sample Performance
7.2.4 Out-of-Sample Performance
7.2.5 Interpretation of Results
7.3 Conclusion
7.3.1 Matlab Program Notes
7.3.2 Suggested Exercises
8 Classification: \ Credit Card Default and Bank Failures
8.1 Credit
Card Risk
8.1.1 The Data
8.1.2 In-Sample Performance
8.1.3 Out-of-Sample Performance
8.1.4 Interpretation of Results
8.2 Banking Intervention
8.2.1 The Data
8.2.2 In-Sample Performance
8.2.3 Out-of-Sample Performance
8.2.4 Interpretation of Results
8.3 Conclusion
8.3.1 Matlab
Program Notes
8.3.2 Suggested Exercises
9 Dimensionality Reduction and Implied Volatility Forecasting
9.1 Hong Kong
9.1.1 The Data
9.1.2 In-Sample Performance
9.1.3 Out-of-Sample Performance
9.2 United States
9.2.1 The Data
9.2.2 In-Sample Performance
9.2.3 Out-of-Sample
Performance
9.3 Conclusion
9.3.1 Matlab Program Notes
9.3.2 Suggested Exercises
Books and book related electronic products are priced in US dollars (USD), euro (EUR), and Great Britain Pounds (GBP). USD prices apply to the Americas and Asia Pacific. EUR prices apply in Europe and the Middle East. GBP prices apply to the UK and all other countries.
Home | Elsevier Sites | Privacy Policy | Terms and Conditions | Feedback | Site Map | A Reed Elsevier Company