Recently, I reread Franses book and expanded my review, which now includes 10 benefits.
(1) Organization by key features of economic time series (trends, seasonality, outliers, conditional heteroskedasticity, non-linearity), rather than by methods, which provides a practical foundation for the various methodologies. The order in which chapters are presented reflects the order of difficulty in modeling trends, seasonality, etc. Even if there were no other benefits, this organization makes it worthwhile.
(2) Appropriate level for first book on time series models as applied to economic time series, explaining more difficult concepts GARCH and VAR without excess detail. Box and Jenksins book is more a textbook; Brockwell and Davis is also more advanced; Hamilton is comprehensive and technical, but not as friendly. This book is very approachable even if you have had only 1 or 2 statistics courses. In economics, many people are interested in forecasting, and Franeses here is a good start. If you are looking for a more advanced forecasting book, try the recent books by Clements and Hendry from Cambridge U Press.
(3) Clear distinction of the steps of model identification, estimation, diagnostics, and selection; something which other time series analysis books do not seem to do early or easily. (4) Delineates stochastic and deterministic models in the second chapter, providing a framework for when to take differences (eg. ARMA vs ARIMA). His timing is excellent. Many people I have interviewed on time series do not understand why they need to difference (eg use prices instead of returns) or why to transform the series (eg use logs instead of actual values).
(5) Generous use of examples with real not simulated data with a website to download all the data, making it possible to import, graph, and analyze on your own.
(6) A website containing printing corrections. Techincal books are likely to have some errors, but very few keep websites to list what those are.
(7) Revealing graphics, especially for conditional heteroskedasticity, the 'CH' in GARCH. Figures 7.1-7.3 illustrate the concept that large returns tend to follow large returns very cleanly.
(8) His notation is clear and consistent, yet not overwhelming: conventional Greek letters, only 1 level of subscripting, matrix noation where appropriate; even the results are neatly presented, as standard errors appear in () below their point estimates. Finally, Franses uses the same notation from chapter to chapter where the term is the same--not so common when chapters written by different authors.
(9) Great appendices: extensive and updated references, a thorough subject index, and an author index. My only suggestion for improvement is that a second edition or the website should contain some exercises. Highly recommended.
(10) The price! There are books published under Wiley at 3 to 4 times the price! under Springer Verlag for 2 to 3 times the price. Certain books are worth the money, but Cambridge University Press paperback publications, when written well, are exeptional values. I encourage the ambitious time series student to look at other time series books, including one written this year by Franses including Quantitative Models in Market Research.