The econometric analysis of seasonal time series. With a foreword by Thomas J. Sargent (Q2768498)

This book deals with the econometric analysis of seasonal time series. It is written for researchers and graduate students familiar with time series analysis at an advanced level, i.e., the reader should have a sound knowledge of recent developments in time series econometrics. Furthermore, some basic knowledge with spectral domain techniques is required.NEWLINENEWLINENEWLINEThe first chapter is a guided tour to the substantive material of the following chapters. The chapter contains a set of examples of seasonal economic and financial time series to motivate the discussion. The examples considered include some empirical series and theoretical models. This survey chapter is followed by a chapter on deterministic seasonality including dummy variable representations and the trigonometric representation. The chapter closes with some test procedures that allow testing for deterministic seasonality. Seasonal unit root processes are the content of chapter three. The authors state the properties of seasonal unit root processes and provide testing procedures for seasonal integration. In addition the chapter contains some extensions such as near seasonal integration as well as higher order nonstationarity.NEWLINENEWLINENEWLINESeasonal adjustment procedures are discussed in chapter four. The adjustment programs considered are the traditional Census X-11, the improved Census X-12 ARIMA, and TRAMO/SEATS. The different adjustment mechanisms are discussed in detail as well as shortcomings of the different approaches and their various options. Nevertheless, the authors avoid a ranking of the procedures so that the reader has to make his own choice about the usefulness of the standard procedures discussed in this chapter.NEWLINENEWLINENEWLINEEstimation and hypothesis testing with unfiltered and filtered data are considered in chapter five. The main theme of this chapter compares the advantages and disadvantages of using unadjusted data as well as seasonally adjusted data. Periodic models of seasonality were introduced into the econometrics literature during the late 1980s. Chapter six reviews these developments. Traditionally, the topic of seasonality is put in a context of macroeconomic time series sampled at a monthly or quarterly frequency. Nevertheless, in financial econometrics new high-frequency data sets are now available with series sampled on a transaction basis; the availability of such series has created considerable interest in models for so-called intra-day seasonality. Procedures to deal with such problems are discussed in chapter 7, together with other topics related to the nonlinear analysis of seasonal time series.