Economic time series, edited by William R. Bell, Scott H. Holan, Tucker S. McElory, Boca Raton, Chapman & Hall / CRC Press, 2012, xvii+535 pp., £63.99 or US$99.95 (hardback), ISBN 978-1-4398-4657-5
Macroeconomic policy formation requires the proper modelling of time series data. Modelling or adjusting for seasonality in economic time series is also important since many time series data feature seasonality. This modelling and interpretation can have far-reaching implications among the policy-making circles. Data providers in different countries around the globe, especially the US Census Bureau, are using various methods to adjust data for seasonality to make available high-quality data for research. The developments in statistical software make handling of seasonality relatively easy for researchers; in this David Findly of the US Census Bureau played a significant role. This book is a collection of 21 articles on economic modelling and seasonality to honour Findly.
The book has 21 chapters contributed by different authors from academia, statistical bureaus and central banks around the world. These 21 chapters are grouped into seven sections based on different themes.
The first three chapters are about periodic modelling of time series. The first chapter discusses the seasonal vector time series model that allows for variations in cyclical components, a multivariate extension of periodic unobserved component (PUC) time series using maximum likelihood. The third chapter by Robert Lund is a comparison of periodic and seasonal ARMA models.
The next section, estimating time series components with misspecified models, deals with STC models and errors in business cycle estimates in seasonal adjustment data. The last chapter in this section by Richard B. Triller deals with seasonal adjustments of US labour force survey data, taking into consideration survey error, using X11/X12 ARIMA and SEATS methods.
The third section explains the quantification of errors in seasonal adjustments including a chapter covering practical problems that arise in seasonal adjustment: developing asymmetric trend-cycle filters, dealing with both temporal and contemporaneous benchmark constraints, detecting trading-day effects in monthly and quarterly time series, and using diagnostics in conjunction with model-based seasonal adjustments.
The fifth section of the book throws light on another important area of research in times series econometrics; handling outlier and extreme values. Galeano and Pena use a modified Bayesian information criterion (BIC) to detect outliers in ARIMA models and this section provides a new way to construct price indexes for financial assets. This section also provides alternative models for seasonal unit root and Bayesian models for long memory time series and for seasonal and calendar effects.
The book provides extensive overview of the various issues concerning seasonal time series modelling with some applications in real-time data. The application parts make the book more interesting to the readers as it helps them to get real-time experience. Overall, the book shares the experience of a set of people from academia and policy-making circles who deal with seasonal time series data extensively, and above all an excellent work to honour David Findly.
Research on time series modelling and applications in real-time data plays an important role in improving the statistical base of different countries and of international organizations. This book covers important research in time series analysis and econometrics, and applications to real-time data are also given. Some chapters of this book provide advanced techniques for time series data analysis. This book can be used by researchers who are doing research in time series analysis and econometrics. Advanced theory on seasonal adjustment given in this book can be used for evaluating whether different aspects of the economy are in growth or decline and given models on seasonal adjustment are enable to draw more reliable inferences about economic activities.