【Abstract】Miller (1977)'s short-sale constraints hypothesis and Merton (1987)'s investor recognition hy-pothesis infer opposite relationships between ownership breadth and future stock returns. We find the mixed empirical evidence in prior literature comes from opposite effects of positive and negative breadth changes on returns. The breadth-future return relationship is positive when breadth decreases, whereas the relationship becomes negative when breadth increases. Our re-sults suggest that the investor recognition hypothesis dominates when breadth increases, whereas the short-sale constraints hypothesis dominates when breadth decreases. This reconciles not only the conflicting evidence but also the predictions of Miller (1977) and Merton (1987).