Analysis of financial time series. 3rd ed. (English) Zbl 1209.91004

Wiley Series in Probability and Statistics. Hoboken, NJ: John Wiley & Sons (ISBN 978-0-470-41435-4/hbk; 978-0-470-64456-0/ebook). xxiii, 677 p. (2010).
The book under review is the third, extended and revised, edition of [Analysis of financial time series. 2nd ed. Wiley Series in Probability and Statistics. Hoboken, NJ (2005; Zbl 1086.91054)] and [Analysis of financial time series. Chichester: Wiley (2002; Zbl 1037.91080)].
In the author’s words, “The primary objective of the revision is to update the data used and to reanalyze the examples so that one can better understand the properties of asset returns. At the same time, we also witness many new developments in financial econometrics and financial software packages. The second goal of the revision is to include R commands and demonstrations, making it possible and easier for readers to reproduce the results shown in the book.
A brief summary of the added material in the third edition is:
1. To update the data used throughout the book.
2. To provide R commands and demonstrations. In some cases, R programs are given.
3. To reanalyze many examples with updated observations.
4. To introduce skew distributions for volatility modeling in Chapter 3.
5. To investigate properties of recent high-frequency trading data and to add applications of nonlinear duration models in Chapter 5.
6. To provide a unified approach to value at risk (VaR) via loss function, to discuss expected shortfall (ES), or equivalently the conditional value at risk (CVaR), and to introduce extremal index for dependence data in Chapter 7.
7. To discuss applications of cointegration to pairs trading in Chapter 8.
8. To study applications of dynamic correlation models in Chapter 10.
I do not include credit risk or operational risk in this revision for three reasons. First, effective methods for assessing credit risk require further study. Second, the data are not widely available. Third, the length of the book is approaching my limit.”
The book covers a compellingly broad range of topics from classical and modern financial econometrics. Chapter 1 discusses the elementary characteristics of financial time series. Chapter 2 introduces linear time series analysis, in particular the AR, MA and ARMA models. Chapter 3 covers some conditional heteroscedastic models, e.g. the ARCH, GARCH, EGARCH, CHARMA models and some stochastic volatility models. Chapter 4 treats the nonlinear analogues of some of the models discussed in the previous sections. Chapter 5 presents some elements of high-frequency data analysis. Chapter 6 is an introduction to continuous-time models e.g. the Black-Scholes model and jump diffusion models. Chapter 7 discusses risk measures, e.g. value at risk and expected shortfall, and presents some elements of extreme value theory. Chapter 8 introduces multivariate time series analysis and presents the vector-valued analogues of the models of the previous sections. Chapter 9 is on principal component analysis and factor models. Chapter 10 presents the multivariate analogues of the models of Chapter 3. Chapter 11 guides the reader through some state-space models and recalls the classical Kalman filter algorithm. Chapter 12 discusses how Markov chain Monte Carlo methods can be applied to some aspects of time series analysis, e.g. linear regression with time series errors, the problem of missing values and outliers, and the parameter estimation of stochastic volatility models. Each chapter is built around examples, and concludes with exercises and references.
The treatment of the material in the book is extremely fast-paced. This may be justifiable, since covering of a wide variety of topics may only be possible if the author omits the mathematical derivation of formulas and rigorous justification of statements. However, an inconvenient side effect of this approach is that the misprints and subtle errors which inevitably appear in such a long work may make it disproportionately hard, especially for a less experienced reader, to follow and correct certain arguments. Also, the reviewer often had difficulties in finding where the terminology and notation of certain subsections were introduced. Nevertheless, all in all the book can be a very useful reference for students as well as for professionals.


91-01 Introductory exposition (textbooks, tutorial papers, etc.) pertaining to game theory, economics, and finance
91G70 Statistical methods; risk measures
62M02 Markov processes: hypothesis testing
62M05 Markov processes: estimation; hidden Markov models
62M07 Non-Markovian processes: hypothesis testing
62M09 Non-Markovian processes: estimation
62M10 Time series, auto-correlation, regression, etc. in statistics (GARCH)
62M20 Inference from stochastic processes and prediction
62P20 Applications of statistics to economics
91B26 Auctions, bargaining, bidding and selling, and other market models
91B30 Risk theory, insurance (MSC2010)
91B70 Stochastic models in economics