Modeling Financial Time Series with S-PLUS[R].

by Eric ZIVOT and Jiahui WANG, New York: Springer-Verlag, 2003, ISBN 0-387-95549-6, xix + 632 pp., $59.95 (softcover).

One probably understands from the size of this book that there is a lot more here than instruction in how to model time series using S-PLUS. That is certainly one component

of the book. The book is a user's guide for S+FinMetrics, a new component of S-PLUS that provides a vast array of statistical functions for analyzing financial time series data. Previously, Lumley (2001) reported a similar book (Millard and Neerchal 2001) for the S-PLUS module for the analysis of environmental data. This is also a fairly comprehensive textbook on the methodology for the statistical analysis of financial time series data.

The book begins with two chapters that introduce S and S-PLUS and their capabilities for specifying and visualizing time series data. Another introductory chapter presents the segmentation of time series methods used throughout the rest of the book: univariate time series, nonstationary time series, long-memory time series, and multivariate time series. There follow five chapters on the first three time series categories, all of which deal with univariate data, where topics include unit root tests, modeling extreme values, time series regression, generalized autoregressive conditional heteroscedasticity (GARCH) modeling, and long memory time series data. The material on univariate time series concludes with a chapter on the rolling analysis of time series. An unfamiliar procedure to me, this gives a methodology for testing a model on historical data to evaluate its stability and predictive accuracy, a concept that BP uses in doing modeling for refinery units.

Next are five chapters that present topics in modeling for multivariate time series. Chapter topics include systems of regression equations, vector autoregressive (VAR) models, cointegration, multivariate GARCH models, and state-space models. A couple of special applications chapters, one on factor models for asset return and the other on the term structure of interest rates, come next. The book concludes with a chapter concerning the use of time series regression models to perform robust tests for the detection of change.

Because the book is about S-PLUS as much as time series analysis, readers will find the text on time series methods somewhat terse. The recent financial series book by Tsay (2002), reported by Ziegel (2002), would be a good companion volume. Good software is a necessary requirement for implementing this methodology. On that issue, this book scores highly. It is absolutely replete with applications and the necessary S-PLUS code.

REFERENCES

Lumley, T. (2001), Review of Environmental Statistics with S-PLUS, by S. Millard and N. Neerchal, Technometrics, 43,495.

Millard, S., and Neerchal, N. (2001), Environmental Statistics with S-PLUS, Boca Raton, FL: CRC Press.

Tsay, R. (2002), Analysis of Financial Time Series, New York: Wiley.

Ziegel, E. (2002), Editor's Report on Analysis of Financial Time Series, by R. Tsay, Technometrics, 44, 408.