For a Gaussian MA(1) process, a new exact ML estimator is proposed that avoids the pile-up phenomenon (boundary estimates). Finite sample comparison is undertaken, along with Wald-type inference for an MA unit root or over-differencing (stationarity).

This paper considers the asymptotic distribution of the OLS estimator in a simple, Gaussian unit-root AR(1) with fixed, non-zero startup. All asymptotic possibilities are considered. The approach is new, relatively simple, and relies on observing and determining the asymptotic/limiting behavior...

Purpose – The paper's objectives are: to address the issue of cointegration (efficient market hypothesis) between Greek spot and futures markets over the period of the crisis, 1999‐2001; to investigate the short‐run and long‐run efficiency of the FTSE/ASE‐20 stock index futures contract...