The Econometric Analysis of Time SeriesThis new edition of A.C. Harvey's clearly written, upper-level text has been revised and several sections have been completely rewritten. There is new material on a number of topics, including unit roots, ARCH, and cointegration. The Econometric Analysis of Time Series focuses on the statistical aspects of model building, with an emphasis on providing an understanding of the main ideas and concepts in econometrics rather than presenting a series of rigorous proofs. It explores the way in which recent advances in time series analysis have affected the development of a theory of dynamic econometrics, sets out an integrated approach to the problems of estimation and testing based on the method of maximum likelihood, and presents a coherent strategy for model selection. A.C. Harvey is Professor of Econometrics at the London School of Economics. |
Contents
Introduction | 1 |
3 | 80 |
Numerical Optimisation | 122 |
Test Procedures and Model Selection | 146 |
Regression Models with Serially Correlated Disturbances | 191 |
Dynamic Models I | 225 |
Stochastic Difference Equations | 264 |
Simultaneous Equation Models | 313 |
Appendix on Matrix Algebra | 359 |
Answers to Selected Exercises | 369 |
380 | |
386 | |
Other editions - View all
Common terms and phrases
alternative applied approach appropriate approximation ARMA assumed assumption asymptotic basis becomes carried coefficients computed condition Consider consistent constructed contains covariance matrix defined denotes dependent derivatives discussion distribution disturbance term dynamic econometric economic effect efficient elements equal equation error example exogenous variables expected explanatory variables expression follows function given gives identifiability implies important independent iterative known large samples leads least squares likelihood likelihood function linear matrix mean method ML estimator multiplier multivariate normally distributed Note null hypothesis observations obtained parameters particular polynomial possible prediction problem procedure properties provides reasonable recursive reduced form regarded regression model residuals respect restrictions result shown significance specification statistic stochastic structure suggests sum of squares Suppose taken taking test statistic theory transformation usually values variance vector written y₁ yields zero