Asymptotic Power Advantages of Long-Horizon Regression Tests

Abstract
Local asymptotic power advantages are available for testing the null hypothesis that the slope coefficient is zero in regressions of y(t+k)-y(t) on x(t) for k > 1 where the x(t) and the change in y(t) are I(0) series. The advantages of these long-horizon regression tests accrue in a linear environment over empirically relevant regions of the admissible parameter space. In Monte Carlo experiments, small sample power advantages to long-horizon regression tests accrue in a region of the parameter space that is larger than that predicted by the asymptotic analysis.

This publication has 24 references indexed in Scilit: