characterization of the bias in case this condition is violated. A feasible bias correction method is shown to virtually eliminate the … bias over a large part of the parameter space …

theory reveals that components from both bias and variance terms contribute to variability in the asymptotics. The errors in … the literature arise because random variability in the bias term has been neglected in earlier research. In stationary … finite sample performance. Implications of the findings for rate efficient estimation are discussed. Simulations in the …