Title:Analysis of Integrated and Cointegrated Time Series with R
Author:Robert Gentleman
Kurt Hornik
Giovanni Parmigiani
Introduce:
This book is divided into three parts. In the first part, theoretical concepts of time series analysis, unit root processes, and cointegration are presented.
Although the book’s aim is not a thorough theoretical exposition of these
methods, this first part serves as a unifying introduction to the notation used
and as a brief refresher of the theoretical underpinnings of the practical examples
in the later chapters. The focus of the second part is the testing of
the unit root hypothesis. The common testing procedure of the augmented
Dickey-Fuller test for detecting the order of integration is considered first. In the later sections, other unit root tests encountered widely in applied econometrics, such as the Phillips-Perron, Elliott-Rothenberg-Stock, Kwiatkowski-Phillips-Schmidt-Shin, and Schmidt-Phillips tests, are presented, as well as the case of seasonal unit roots and processes that are contaminated by structural shifts. The topic of the third and last part is cointegration. As an introduction, the two-step method of Engle and Granger and the method proposed by Phillips and Ouliaris are discussed before finally Johansen’s method is presented.The book ends with an exposition of vector error-correction models that are affected by a one-time structural shift.