8

339-404

66

 

John Wiley & Sons, Inc. (Visit publisher website)

2005

English

 

 

 

Advanced

 

 

Quantitative analysis, multivariate time series analysis, weak stationarity, cross-correlation matrices, linear dependence, multivariate Portmanteau tests, vector autoregressive (VAR) models, stationarity condition and moments of a VAR(1) model, vector AR(p) models, building a VAR(p) model, impulse response function, vector moving-average models, vector ARMA models, unit-root nonstationarity and cointegration, cointegrated VAR models, forecasting of cointegrated VAR models, threshold cointegration and arbitrage, multivariate threshold model, matrix operations, matrix inverse, trace, eigenvalue, eigenvector, positive-definite matrix, multivariate normal distribution

 

This chapter is useful for researchers and investment managers seeking to identify relationships between related time series. The author emphasizes intuition relevant for application rather than theoretical development. Interpretations of fitted models provide useful guides to implementation for other series. Model properties, methods of model selection and estimation are presented for each model contained in the chapter. Using estimated models to forecast and produce forecast standard errors is illustrated in most cases. Useful S-Plus examples are provided throughout.

 

8.1 Weak Stationarity and Cross-Correlation Matrices
8.1.1 Cross-Correlation Matrices
8.1.2 Linear Dependence
8.1.3 Sample Cross-Correlation Matrices
8.1.4 Multivariate Portmanteau Tests
8.2 Vector Autoregressive Models
8.2.1 Reduced and Structural Forms
8.2.2 Stationarity Condition and Moments of a VAR(1) Model
8.2.3 Vector AR(p) Models
8.2.4 Building a VAR(p) Model
8.2.5 Impulse Response Function
8.3 Vector Moving-Average Models
8.4 Vector ARMA Models
8.4.1 Marginal Models of Components
8.5 Unit-Root Nonstationarity and Cointegration
8.5.1 An Error-Correction Form
8.6 Cointegrated VAR Models
8.6.1 Specification of the Deterministic Function
8.6.2 Maximum Likelihood Estimation
8.6.3 A Cointegration Test
8.6.4 Forecasting of Cointegrated VAR Models
8.6.5 An Example
8.7 Threshold Cointegration and Arbitrage
8.7.1 Multivariate Threshold Model
8.7.2 The Data
8.7.3 Estimation
Appendix A: Review of Vectors and Matrices
Appendix B: Multivariate Normal Distributions
Appendix C: Some SCA Commands
Exercises
References

 

If you are interested in purchasing the book, please click here.

 

The Second Edition of this critically acclaimed text provides a comprehensive and systematic introduction to financial econometric models and their applications in modeling and predicting financial time series data. This edition continues to emphasize empirical financial data and focuses on real-world examples. Following this approach, readers will master key aspects of financial time series, including volatility modeling, neural network applications, market microstructure and high-frequency financial data, continuous-time models and Ito`s Lemma, Value at Risk, multiple returns analysis, financial factor models, and econometric modeling via computation-intensive methods.

The author begins with the basic characteristics of financial time series data, setting the foundation for the three main topics:

* Analysis and application of univariate financial time series
* Return series of multiple assets
* Bayesian inference in finance methods

This new edition is a thoroughly revised and updated text, including the addition of S-PlusŪ commands and illustrations. Among the new material, readers will find:

* Consistent covariance estimation under heteroscedasticity and serial correlation
* Alternative approaches to volatility modeling
* Financial factor models
* State-space models
* Kalman filtering
* Estimation of stochastic diffusion models

This is an ideal textbook for MBA students as well as a reference for researchers and professionals in business and finance.