A Review of Unit Root Tests in Time Series: Volumes 1 and 2

Abstract

This is a two‐volume synthesis by Kerry Patterson of the voluminous unit root literature. In the Preface to Volume 1, the author states that ‘The purpose of this book is to provide a review and critical assessment of some key procedures related in one way or another to the problem of testing for a unit root in a stochastic process’. Overall, it is my opinion that this two‐volume set achieves these goals. Both volumes are carefully written and cover a huge amount of ground, as you might expect given that together they total some 1191 pages of text. Very often text books in this area simply summarise the contents of a huge number of papers without providing any critical insight, additional contributions or indeed coherent overview of the topics involved. That is not the case with these two volumes. The author includes his own Monte Carlo work throughout both volumes to investigate, explore and assess further the usefulness of the procedures he discusses. Careful empirical illustrations are also included throughout to help develop the underlying ideas and communicate these better to an applied audience. I also found a genuine attempt by the author to link topics together in a meaningful and insightful way. Practical issues that affect practitioners in applying procedures to real data are discussed in detail, rather than simply ignored as they often have been in the original articles. This book provides an excellent guide for applied researchers (both in academic and academic‐related professions) and for masters and doctoral level students undertaking applied research on macroeconomic and financial data. Each chapter contains a very useful set of exercises designed to help the reader look more deeply into the topics discussed.