Time Series and Dynamic Models

Christian Gourieroux is a leading researcher in time series analysis. This text, which he penned with Alain Monfort, was translated from the original French by Giampiero Gallo. In my opinion, it is the best general text on time series analysis available. It is a masterpiece.

 

Organization is impeccable. Results flow seamlessly from one to the next. The writing, for the most part, is very accessible. It nicely balances mathematical formality with a hands-on, tone that tells you what is "really going on." Discussions of how to fit models to data are especially user-friendly. Proofs are offered if they are short and enlightening. Otherwise, a reference is cited, and the authors move on. There are plenty of examples that lend real insights. One is of a pathological white noise that will challenge your perception of white noise. If you were to see a realization of this white noise, you would be sure it could not be from a white noise, but you would be mistaken! Another nice feature is that, unlike many books on time series analysis, this one has exercises.

Having said all this, I should caution that this may not be the right book for you. There are two reasons.

The first is that the book employs some high level mathematics. This is one reason why it is as good as it is. Important results are achieved with minimal effort because of the math the authors employ. By comparison, Hamilton (1994) is often tedious and difficult to follow because of the less powerful mathematics it uses. Do you know what it means for two functions to be orthogonal? Have you ever worked with infinite-dimensional vector spaces? Have you taken a graduate level course in probability? If you answer "no" to these questions, you will probably find the book is over your head. If you answer "yes", then you are likely to enjoy the style of this book.

Contents

Traditional Methods

1. Introduction

2. Seasonal adjustment by regression methods

3. Moving averages for seasonal adjustment

4. Exponential smoothing methods

Probabilistic and Statistical Properties of Stationary Processes

5. Some results on the univariate processes

6. The Box and Jenkins approach to forecasting

7. Multivariate time series

8. Time series representations

9. Estimation and testing (stationary case)

Time Series Econometrics: Stationary and Nonstationary Models

10. Causality, exogeneity and shocks

11. Trend components

12. Expectations

13. Specification analysis

14. Non-stationary processes

State Space Models

15. State space models and the Kalman filter

16. Applications of the state space model

Another reason why the book may not be right for you is the choice of topics. This is a book primarily related to econometrics—the application of time series analysis to economic problems. Topics with little relevance to finance—such as smoothing out seasonality—are covered in detail. Some topics that are highly relevant to finance—such as ARCH and GARCH models of conditional heteroskedasticity—are not covered. This need not be a show-stopper. Even if your primary interest is in finance, there is a tremendous amount of important information covered here. This includes excellent discussions of MA, AR, ARMA and state-space models.

If you are interested in time series analysis—and you have the necessary math skills—I strongly encourage you to read this book. If you have not already done so, begin with a more basic introduction, such as Franses (1998) or Harvey (1993), but then rapidly proceed to this book.

 

 

Ads by Contingency Analysis.

Advertise on this site.
 

disclaimer

website: http://www.contingencyanalysis.com
books direct link: http://www.riskbook.com
copyright © Contingency Analysis, 1996 - current