Mean Reversion in Stock Prices: An Error‐Correction Approach

Several recent studies have indicated the existence of a predictable component in stock prices. This study examines the sources of this serial correlation using error‐correction models. The results show that autocorrelated economic variables can generate serial correlation in stock returns. After these effects are accounted for, however, significant serial correlation in stock prices remains. The activities of noise traders and inefficiencies in the pricing of securities, within the context of limitations to the arbitrage process, are suggested as additional sources of serial correlation in stock prices.