MEAN REVERSION ON SINGAPORE STOCK PRICES

Abstract

Since the work of Fama and French (1988a) and Poterba and Summers (1988), there has been growing controversy among economists on whether stock returns are mean reverting. This paper employs a new family of spectral shape tests, proposed by Durlauf (1991), and variance ratio methodology, developed by Lo and MacKinlay (1988), to test the behavior of stock price on the Singapore Stock Exchange (SES). The data set consists of weekly returns from 1980 to mid 1995. Our results suggest that stock prices in Singapore have not followed a random walk in the past 15 years. For short investment horizon, we find strong evidence of positive autocorrelation, with variance ratios above one; for longer horizons, we find indications of negative autocorrelation, with variance ratios below one. This phenomenon of positive serial autocorrelation in short run and negative autocorrelation in the long run is called mean reversion. Despite the low power at traditional significant levels, the patterns of mean reversion are similar to the results of Poterba and Summers on the U.S. stock market. Our findings may have important implications for both academics and practitioners.