Customer Reviews

Analysis of Financial Time Series

5 star:		(8)
4 star:		(3)
3 star:		(1)
2 star:		(2)
1 star:		(0)

 

 

Written by a University of Chicago professor, this book comprehensively covers times series topics relative to investment and trading-oriented finance (i.e., Wall Street money-making machines). Treatment is generally clear and thorough, but an advanced math and stat background is an absolute prerequisite for understanding the materials.

S-Plus/R code is given, but strangely, there is very little on *why* and
*when* one uses each of the techniques. Under what cirmcustances should I use or not use GARCH? What exactly is PCA good for in real-world applications? These important questions are not answered, in other words, you don't get a sense of the real-world context for these topics.

Professor Tsay is a student of the Wisconsin school of statisticians where he learned time series from Box and Tiao. He is an excellent lecturer and a good writer. I have attended one of the short courses he taught on time series. New models have been developed to deal with the special behavior of financial time series. Professor Tsay is always at the forefront of that research and teaches at Chicago in one of this country's top business schools. If I am correct George Tiao is also there at present.

This is the second edition of a popular text. Financial time series play an ever more important role in our lives during these turbulant economic times. Tsay cover the tradition Box-Jenkins models but these models are not always appropriate for financial data. So he also introduces the GARCH models and some nonlinear models. The book includes some models that I am not familiar with. I have done research in time series but never with financial data. There is some theory involving stochastic differential equations that explains some of the turbulant behavior of financial series. The text by J. Michael Steele provides thorough coverage to this theory.

Tsay also deals with the pesky problem of outliers. A very practical problem that is often ignored in other econometric texts. He also has a chapter on Bayesian approaches. Some computing in SPlus is also included in this revision of the text.

This is not a reference book, and it's not about "big" theory either. It's pretty practical, and good for self study. You should have access to some econometric/statistical software (i.e. EViews, S-Plus, etc.) to fully understand this book.

I had a detailed study of the whole book before finally deciding to buy in on web. As a statistician and a beginner on Math Finance, I would say this book deserves every penny I spent on it.

The author's intention to make it a reference book can be appreciated by both educators and practitioners. It starts with a couple of chapters on the ARIMA and the GARCH models. Little theoretic depth was explored yet the algorithms and the procedures for solution are emphasized. After that, the topic switches to the nonlinear time series modeling and high-freq data analysis. This part is, and will be, rather confusing to readers with less training in financial economics and theories are reluctantly clearly stated. What follows is a single chapter of so-called continuous time models and it is actually a sketch of the first few chapters of any mathematical finance textbook. Literally, this chapter is all about Black-Scholes and a little jump-diffusion model. The major reason why I called it a reference book is because it includes one chapter on VaR between the math finance chapter and the multi-variate time series models part. The author didn't say much more than that VaR is essentially some quantile calculation, which is fine in the statistical meaning. However, this description seems really "shallow" as compared with Jorion's book on VaR and risk management.

After all, I would give it a five star because its comprehensiveness and the author's effort to incorporate so many things in order to re-define the framework of financial time series analysis.

This book is an excellent toolbox for anyove dealing in the field of financial engineering, however, as a real toolbox, the author doesn't explain the exact use of all tools and how to interpret the results. This is why this book is for advanced users who need a well documented reference but it is not very suitable for beginners in the field. The Splus code is welcome.

This book is awesome. It starts with bedrock concepts needed for analysis of financial data and it takes the student up to the most recent and important techniques used in the industry today. However, if one expects to fully utilize this text, one should have at least one semester of applied econometrics or some equivalent course in statistics and continuous probability, although it will be practical to study the two topics concurrently.

This text is absolutely perfect for Masters students learning financial econometrics. There is a little theory, clear explanations, and quite a few real world examples. (I don't think any text would tell the reader what model to use when, because that's application-specific.) It assumes some knowledge of finance and basic econometrics/statistics, which is fair enough. To get more theory, Hamilton (1994) remains the authority, and Campbell, Lo, MacKinlay (1997) is a great introduction for PhD students, and generally an ideal companion volume to this one.

The coverage of the topic is broad and deep. It is one of the few introductory books that devotes some space to transfer function modeling and does so intelligibly.
A must have for the novice as well as those more familiar with the topic that need a solid reference.

First of all, it is well written in a very practical point of view. The whole book is aimed fullly to real financial data(appended in the author's web). People can gain not only the well-explained theories but the hand-on experience with data analysis using SPLUS or any other package.
Secondly, the author is a real expert in this field and has been publishing lots of nice work. All models in the book are clearly illustrated and commented.
Thirdly, it covers a lot of topics in analysis of FT. Reader can learn almost all the valuable things in this field from this book.

If anyone wanna truly learn this book, she/he has to sit down and plays some real data on computer. I think this is the best way and the only way to use this book.

I am new to TS analysis and I bought this book because I am always interested in how statistics works in finance. I can say this is probably the best book for self-study. The writting is clear and the examples are both interesting and easy to follow.

Interestingly the other day I asked my friend, who has a Ph.D. in statistics, to recommend a book in time series, and he immediately produced this book from his bag and said this is a very good book.