Efficient tests of stock return predictability
Tests of the predictability of stock returns may be invalid when the predictor variable is persistent and its innovations are highly correlated with returns. This paper proposes two methods to deal with the problem. First,we develop a pretest that determines when the conventional t-test is misleading. Second,we develop a new test of predictability that always leads to correct inference and is efficient compared to existing methods. Applying our methods to US data,we find that the conventional t-test is highly misleading for the dividend-price ratio and the smoothed earnings-price ratio. However,we find evidence for predictability using our test, particularly with the earnings-price ratio. We also find evidence for predictability with the short-term interest rate and the long-short yield spread,for which the conventional t-test leads to correct inference.
For young readers
Book data: Open Library