Power and size properties of alternative test of autocorrelation in dynamic models

Abstract

The four most readily available tests of autocorrelation in dynamic models namely Durbin’s M test, Durbin’s H test, Bruesch Godfrey test and Q test are compared in terms of their power and size for varying sample sizes, levels of autocorrelation and level of significance using Monte Carlo simulations. Ten thousand simulations in STATA reveal that Durbin M test is the most encompassing and generally performs the best in all sample sizes. Bruesch Godfrey has comparable and at times minutely better performance than Durbin’s M test however in small sample sizes, Durbin’s M test outperforms the Bruesch Godfrey test in terms of power. Durbin’s H has the drawback of being inapplicable in many cases where as Q test’s consistently worst performance can be attributed to the wrong specification of error.