Time Series and Dynamic Models (Themes in Modern Econometrics)

Rating:
Summary: Some topics in time series
Review: the book's goal are nested test, ARIMA models, and the book developos the clasical basic theory of seasonality and Kalman Filter but the book needs Intermediate mathematics and advanced statistics, the book is good as reference in topics of econometrics but if you needs learn time series the Hamilton's book is a good idea.

Rating:
Summary: excellent reference book
Review: This book is a graduate level introduction to time series econometrics. By introduction I mean that it covers a large amount of material, from stationary ARMA models to cointegration, without going too deep into details, although it requires strong mathematical and statistical background. The aim of this book was to be a source of references for applied researches and a textbook for graduate level courses. So in this respect it's similar to Hamilton's "Time Series Analysis". I think that as a textbook Hamilton's book works better, because the derivation of many results in it is more detailed. On the other hand, Gourieroux and Monfort treatment of many topics is more advanced. Gourieroux and Monfort provide much more information, for example, on uses of frequency domain. They also included a chapter on fractional processes.

Rating:
Summary: Don't Buy It
Review: This is a crap book. Don't buy it. I have 3 books by this author, all published by CUP, and I swear I will never and ever buy his fourth book. This is the worst book I've every read. There are several common weak points of his books -- confusing symbols, lack of explanations on those necessary issues and lengthy B.S. on those simple issues. Numerous typing errors make the matter worse. This book spend a chapter talking about old fashion of moving averages, such as Spencer 7-point and 15-point. It almost goes into the field of graduation. What's the point? The chapters on ARIMA are also rubbish. My feeling is that the author lacks sense in statistics, all he saw are just mathematics. On the whole, this book is just on the wrong field, at the wrong time and with the wrong title.

Rating:
Summary: Don't Buy It
Review: This is a crap book. Don't buy it. I have 3 books by this author, all published by CUP, and I swear I will never and ever buy his fourth book. This is the worst book I've every read. There are several common weak points of his books -- confusing symbols, lack of explanations on those necessary issues and lengthy B.S. on those simple issues. Numerous typing errors make the matter worse. This book spend a chapter talking about old fashion of moving averages, such as Spencer 7-point and 15-point. It almost goes into the field of graduation. What's the point? The chapters on ARIMA are also rubbish. My feeling is that the author lacks sense in statistics, all he saw are just mathematics. On the whole, this book is just on the wrong field, at the wrong time and with the wrong title.

Rating:
Summary: Not a book for economists; particularly hard to understand
Review: When you study an econometrics PhD. In France, you can't avoid reading books of Gourieroux: it's the national hero of econometrics. In fact, when you discover all the research he has done, you have to admit it's a brilliant person. But his books aren't particularly clear. They are made for mathematicians (the only students in my classroom that appreciate this book aren't economists; the rest of us, simple mortals, use more friendly books, such as Davidson and Mackinnon, Greene, Enders and Hamilton). What can I say? It's a very complete book: seasonality is deeply treated, ARIMA models are studied profoundly and, you can even find a spectral analysis chapter and another of the Kalman filter. But they are pretty hard to understand. The notation is complex, more than necessary. There are lots of equations and little explanations. If you are a mathematician, this book will satisfy your needs; it's rigorous and fairly complete (even if the selection of topics it's not the ideal one, I think); if you are not, you should better go to Hamilton's manual. If what you want is a cookbook of time series, then buy Enders.