Modeling Financial Time Series with S-PLUS®

Just to be clear: buying this book does not mean you are buying S+Finmetrics. You need to purchase Splus base + the Finmetrics module separately. Unfortunately I tried to call SPLUS (twice) to obtain an academic license, and no one ever called me back. I ended up getting a copy from my university. I wish SPLUS would set up an online download, where I can simply pay with a credit card and download the product immediately, instead of dealing with sales people. That's a very archaic distribution system in my opinion. But this review is about this book. In fact, this book is AMAZING. It is basically a unique combination of a S+Finmetrics userguide and a primer on financial econometrics. It covers virtually all aspects of modern financial econometrics with an emphasis on practical examples. Theory is discussed to illustrate and motivate the examples. There are no proofs. If you want understand, say, a Vector Autoregression foreasting error decomposition, are you going to slog through Hamilton's "Time Series Analysis" and try to implement it on your own? No, you are going to turn to the nice tidy description in Ch11 of this book, and then call the "fevd" method, so you know what is doing and how to interpret the results. A note on R vs. S+Finmetrics: much of the functionality in S+ Finmetrics is available in R, it's just spread across a lot of different packages. The advantage of a commercial product such as S+ Finmetrics is that it consolidates these packages, and provides (more or less) standardized methods and classes to support them. For example, in R it is possible to fit a long memory ARIMA model using the function fracdiff. However in R the function fracdiff does not return residuals, the inclusion of exogenous x variables or support forecasting (no predict method). In SPLUS, the same function (FARIMA) returns all of these.

As other reviewers have mentioned, this book is useless without FinMetrics. It is merely a user manual for that package, and has hardly any intrinsic value on its own.

This is an excellent book on financial econometrics, very practical yet rigorous. I wish all econometrics/statistics textbook could like this. Basic theory followed by practical examples - real life examples, not simplified ones like in other books. The authors gave detailed instructions on how to implement various econometric models, i.e. multi-factor models, GARCH, MGARCH, long memory models, state-space, etc. Most econometrics textbooks are at two extremes, they are either too theoretical (you still don't know how to put those models in real life), or too simple (lack of mathematical rigor and without advanced applications). This book is a combination of both worlds, computer codes/math models, and real life examples (some really good ones). A lot of cutting-edge techniques and advanced topics are also covered.

The best thing about this book is that it combines financial time series analysis with "real-life" examples that are either reproducible or easily adaptable. Being that it is also the user manual for the S+FinMetrics module for the SPLUS stats. package it also reads like a software manual (some people like that). This book provides a good sample of many time series techniques that can be applied out of the box. Note: The S+FinMetrics module includes this book.