Patterson K. Unit Root Tests in Time Series. Vol 1. 2011 torrent

Textbook in PDF format

Volume 1: Key Concepts and Problems
Testing for a unit root is now an essential part of time series analysis. Indeed no time series study in economics, and other disciplines that use time series observations, can ignore the crucial issue of nonstationarity caused by a unit root.
However, the literature on the topic is large and often technical, making it difficult to understand the key practical issues. This volume provides an accessible introduction and a critical overview of tests for a unit root in time series, with extensive practical examples and illustrations using simulation analysis. It presents the concepts that enable the reader to understand the theoretical background and importance of random walks and Brownian motion to the development of unit root tests. The book also examines the latest developments and practical concerns in unit root testing. This book is indispensable reading for all interested in econometrics, time series econometrics, applied econometrics and applied statistics.
Detailed Contents
List of Tables
List of Figures
Symbols and Abbreviations
Preface
Introduction to Random Walks and Brownian Motion
Why Distinguish Between Trend Stationary and Difference Stationary Processes?
An Introduction to ARMA models
Bias and Bias Reduction in AR Models
Confidence Intervals in AR Models
Dickey-Fuller and Related Tests
Improving the Power of Unit Root Tests
Bootstrap Unit Root Tests
Lag Selection and Multiple Tests
Testing for Two (or More) Unit Roots
Tests with Stationarity as the Null Hypothesis
Combining Tests and Constructing Confidence Intervals
Unit Root Tests for Seasonal Data
Appendix A: Random Variables; Order Notation
Appendix B: The Lag Operator and Lag Polynomials
References
Author Index
Subject Index