An analyst estimates a simple linear regression of a time series $y_t$ on another time series $x_t$. In which of the following situations could the estimates of the regression test statistics be valid?