We present a novel approach to analyzing stock return predictability that accommodates (i) arbitrary predictor persistence, (ii) panels with common factors, (iii) multiple predictors, (iv) short- and long-horizon analysis, and relies on standard inference from least-squares estimation of a...