Time Series Analysis (英語) ハードカバー – 1994/1/11
Kindle 端末は必要ありません。無料 Kindle アプリのいずれかをダウンロードすると、スマートフォン、タブレットPCで Kindle 本をお読みいただけます。
The last decade has brought dramatic changes in the way that researchers analyze economic and financial time series. This book synthesizes these recent advances and makes them accessible to first-year graduate students. James Hamilton provides the first adequate text-book treatments of important innovations such as vector autoregressions, generalized method of moments, the economic and statistical consequences of unit roots, time-varying variances, and nonlinear time series models. In addition, he presents basic tools for analyzing dynamic systems (including linear representations, autocovariance generating functions, spectral analysis, and the Kalman filter) in a way that integrates economic theory with the practical difficulties of analyzing and interpreting real-world data. Time Series Analysis fills an important need for a textbook that integrates economic theory, econometrics, and new results.
The book is intended to provide students and researchers with a self-contained survey of time series analysis. It starts from first principles and should be readily accessible to any beginning graduate student, while it is also intended to serve as a reference book for researchers.
"I am extremely enthusiastic about this book. I think it will quickly become a classic. Like Sargent's and Varian's texts, it will be a centerpiece of the core cirriculum for graduate students."--John H. Cochrane, University of Chicago
A carefully prepared and well written book. . . . Without doubt, it can be recommended as a very valuable encyclopedia and textbook for a reader who is looking for a mainly theoretical textbook which combines traditional time series analysis with a review of recent research areas.--Journal of Economics
This is a great book. Given that it has 799 pages, you must expect a lot of detail, and none of it is fluff. Not only are the procedures for constructing every kind of time series spelled out completely, but several times the author points out potential pitfalls and gives tips and tricks for circumventing them. One of them worked for me in another context and meant the difference success and failure in that project. Another benefit of the abundant detail is that, while there are recipes for each time series type, they are not written as a series of steps, but in paragraphs of detailed text. The result is you tend to understand the material, rather than just mindlessly carrying out a series of instructions. People have performed near miracles with maximum likelihood estimators, and this book tells you how it is done.
Obviously, the book is long, but another Amazon reviewer wrote that he knew exactly what kind of time series he needed, found the instructions to build it in the text, and was done in a day. Because the book has been carefully divided into chapters, sections, and sub-sections, all with clear titles and sub-titles, it is relatively quick and easy to find something, if you know what you need.
There are more recent books for sale at Amazon that claim to contain the results of the latest research on multivariate time series. While this book contains material on multivariate problems, it is presented only as an extension of single-variable situations (in what I have read; I have not finished the book). Since it is hard to avoid having several variables in a complex time series, you may want to consider the newer material.
This is the first time I’ve read a textbook so thoroughly and even solved every single problem after each chapter since college. I read it on numerous subway journeys to home, to school and to office, standing mostly. I read it in the beloved Old Hall of Tsinghua Library, during class breaks at Wudaokou, and at my office desk when I’ve done my work as a central banker. I read it late into the night, when my family all fell asleep, only the dim light from desk lamp as my sole companion. Hardcover Hamilton became softcover and covered by adhesive tape. A white-turned-grey Hamilton of 799 pages and a solution manual of 63 pages are the by-products. Although the manual contains many errors and some proofs are not as simple as appendix, when looking at it, as well as the book itself, it feels amazing. I’ve finally done it.
Some thought as Ph.D., we should read papers instead of textbooks. It’s also ‘boring’ and sufficiently daunting to read those monographs. But as a newcomer to economic theory, considering my background of both math/physics and finance, I choose to start my career as an economist by reading classics. Every Ph.D. should be responsible for his own training. After reading several classic text book written by economics gurus, I’m so glad that I’ve made the right decision.
Hamilton is not only about time series, but also major areas of econometrics. I found it much more superior than any book of econometrics I’ve read. It covers maximum likelihood estimation, asymptotic theory, general least squares, VAR, Bayesian Analysis, General Method of Moments, Cointegration, ARCH, GARCH, IGARCH and many other general topics covered in advanced econometrics courses.
It’s cogent, coherent, rigorous, and most importantly, beautiful. I can’t talk the beauty of Hamilton, but I can name several important chapters. First several chapters are easy and pieces of cakes. Chapter 5 shows abundant numerical optimization techniques, which will blow up your mind for the first time. Chapter 7 is about asymptotic theory. This is the heart of advanced econometrics and repeatedly referenced to through the book. Chapter 8 instilled a whole semester of Advanced Econometrics I which we took last year into 28 pages. These two chapters are the next major blow-your-mind point. Chapter 13 (Kalman Filter) is the first major obstacle readers might encounter. Chapter 17 and 18 cover asymptotic theory for nonstationary time series. Chapter 17 and 19 are not only long, but also freaking difficult. Chapter 20 wraps up nonstationary time series. I find math proof in it truly splendid. Chapter 21 and 22 are the last chapters and written like poems, or musical notes. Yes, sipping through ARCH, GARCH, IGARCH, EGARCH is turned into poem-reading by Mr. Hamilton. I thank him very much for this. For so many years, when I heard about any-ARCH, I frowned. Now I’m more than happy to hear the ARCH family.
Hamilton is hard. Reading speed diverged much during last 6 months. I could finish tens of pages per day, but most of the time, only several pages per day. When reading Chapter 19, I found it so hard that I forgot what the just turned page told. In Chinese, we call it ‘Duanpian('')’. For most of the chapters, I must read more than 3 times to gain a basic understanding. I read a little bit slowly not because Mr. Hamilton is a bad writer, but because the content itself. If you have read Greene’s Econometrics Analysis, you’ll find Hamilton more Ph.D.-friendly.
Once when I was asked about what books to choose for the entrance exam of Ph.D. of PBC School of Finance in Tsinghua University, I would tell them several econometrics textbooks written in Chinese, such as CHENG Qiang’s or JIN Yunhui’s. Now I will definitely recommend Hamilton.
Time Series Analysis by James D. Hamilton is simply the green card to econometrics.