Asymptotic theory for the durbin-watson statistic under long-memory dependence

In time series regression models with "short-memory" residual processes, the Durbin-Watson statistic (DW) has been used for the problem of testing for independence of the residuals. In this paper we elucidate the asymptotics of DW for "long-memory" residual processes. A standardized Durbin-Watson statistic (SDW) is proposed. Then we derive the asymptotic distributions of SDW under both the null and local alternative hypotheses. Based on this result we evaluate the local power of SDW. Numerical studies for DW and SDW are given.