Small sample estimation of a cointegrating vector: an empirical evaluation of six estimation techniques

Abstract

A large number of techniques are now available for estimating a cointegrating regression. Although many of these techniques provide asymptotically equivalent estimators, their small-sample properties are known only with respect to a limited number of Monte Carlo studies. In light of the growing controversy over the nature of non-stationarity of economic time series, a comprehensive evaluation of these techniques within an applied framework can shed more light on the relative merits of these techniques. An estimation of long-run demand elasticities by six such techniques based on annual data from Canada, China and Singapore show rather disconcerting results. In small samples OLS may still be the best choice.