Likelihood-based inference for cointegration of some nonstationary time series

2.1 Introduction

We consider likelihood inference for a class of nonstationary time series and to motivate the discussion we consider some simple economic applications and show how the analysis of the statistical model helps in gaining insight and understanding of economic phenomena. This section contains the basic definitions of integrated variables, cointegration and common trends, and discusses by means of examples the formulation of models and processes in terms of common trends in the moving average representation and of disequilibrium errors in the error correction form of an autoregressive model.