Analysis of Integrated and Cointegrated Time Series with R

verfasst von: Dr. Bernhard Pfaff

Verlag: Springer New York

Buchreihe : Use R!

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

The analysis of integrated and co-integrated time series can be considered as the main methodology employed in applied econometrics. This book not only introduces the reader to this topic but enables him to conduct the various unit root tests and co-integration methods on his own by utilizing the free statistical programming environment R. The book encompasses seasonal unit roots, fractional integration, coping with structural breaks, and multivariate time series models. The book is enriched by numerous programming examples to artificial and real data so that it is ideally suited as an accompanying text book to computer lab classes.

The second edition adds a discussion of vector auto-regressive, structural vector auto-regressive, and structural vector error-correction models. To analyze the interactions between the investigated variables, further impulse response function and forecast error variance decompositions are introduced as well as forecasting. The author explains how these model types relate to each other.

Inhaltsverzeichnis

Frontmatter

Theoretical Concepts

1. Univariate Analysis of Stationary Time Series

Although this book has integration and cointegration analysis as its theme, it is nevertheless a necessity to first introduce some concepts of stochastic processes as well as the stationary ARMA model class. Having paved this route, the next steps (i.e., the introduction of non-stationary, unit root, and long-memory processes) will follow in Chapter 3.

2. Multivariate Analysis of Stationary Time Series

This is the second chapter that presents models confined to stationary time series, but now in the context of multivariate analysis. Vector autoregressive models and structural vector autoregressive models are introduced. The analytical tools of impulse response functions, forecast error variance decomposition, and Granger causality, as well as forecasting and diagnostic tests, are outlined. As will be shown later, these concepts can be applied to cointegrated systems, too.

3. Non-stationary Time Series

In this chapter, models for non-stationary time series are introduced. Before the characteristics of unit processes are presented, the differences between trend- and difference-stationary models are outlined. In the last section, long-memory processes (i.e., fractionally integrated processes) are presented as a bridge between stationary and unit root processes.

4. Cointegration

In the previous chapters, a brief explanation of univariate and multivariate time series models and their characteristics was presented. The focus of this chapter is on the simultaneous modeling of time series and inferences of the relationships between them if some or all of them are integrated processes of order one. As will be shown, the degree of integration and a careful examination of the data-generating processes are of utmost importance. We will begin by briefly reviewing the case of a spurious regression before we proceed by providing a definition of cointegration and its error-correction representation. In the last section, the more encompassing vector error-correction models are presented.

Unit Root Tests

5. Testing for the Order of Integration

This chapter is the first in which the theoretical aspects laid out in Part I of the book are put into “practice.” We begin by introducing the most commonly employed unit root tests in econometrics: the Dickey-Fuller test and its extensions. To discriminate between trend- and difference-stationary time series processes, a sequential testing strategy is described. Other unit root tests encountered in applied research are presented in the ensuing sections.

6. Further Considerations

In Chapter 5, various unit root tests were introduced and compared with each other. This chapter deals with two further topics. First, the case of structural breaks in a time series and how this affects the inference about the degree of integratedness is considered. Second, the issue of seasonal unit roots is discussed, as it was only briefly touched on in Section 3.2.

Cointegration

7. Single-Equation Methods

This is the first chapter of the third and last part of this book. The cointegration methodology is first presented for the case of single-equation models. The Engle-Granger two-step procedure is demonstrated by estimating a consumption function and its error-correction form for the United Kingdom as in Holden and Perman [1994]. In the Section 7.2, the method proposed by Phillips and Ouliaris [1990] is applied to the same data set. The application and inferences of a vector error-correction model are saved for Chapter 8.

8. Multiple-Equation Methods

In this chapter, the powerful tool of likelihood-based inference in cointegrated vector autoregressive models (VECMs) is discussed. In the first section, the specification and assumptions of a VECM are introduced. In the following sections, the problems of determining the cointegration rank, testing for weak exogenity, and testing of various restrictions placed on the cointegrating vectors are discussed. The topic of VECMs that are contaminated by a one-time structural shift and how this kind of model can be estimated are presented. This chapter concludes with an exposition of structural vector error-correction models.

Backmatter

Titel: Analysis of Integrated and Cointegrated Time Series with R
verfasst von: Dr. Bernhard Pfaff
Verlag: Springer New York
Electronic ISBN: 978-0-387-75967-8
Print ISBN: 978-0-387-75966-1
DOI: https://doi.org/10.1007/978-0-387-75967-8