摘要: The slope coefficient estimator in predictive regressions for stock returns is biased by a lagged stochastic regressor. There also spurious regression if the underlying expected return highly persistent. This paper studies how interactions between two biases affect inferences about predictability international equity market returns. analysis considers work presence of data mining variables. I find that can reinforce or offset each other, depending on parameters model. present new correction bias both effects and evaluate its economic significance. Adjusting associated with effects, many global predictors have weak explanatory power when they are individually regressed against world Using local predictors, we cannot reject null hypothesis an outcome 10 18 national once account apparent number searches.