Excessive stock price dispersion: a regression test of cross-sectional volatility

In this paper we apply a regression test of the volatility of asset prices to a cross-section data set of US stock prices each year between 1932-71. We show that the rejection of REEM in the time series domain carries over to a data set consisting of observations on a cross-section of individual share prices within a particular year, and we refer to this phenomena as excess dispersion of stock prices. In nearly all of the years over the period 1932-1971 we find that stock prices are excessively dispersed. This finding is consistent with the existence of a firm specific bubble which drives a wedge between the values of pt* and pt. We go on to examine the relationship between the mis-pricing and market fundamentals which we take to be related to past dividends. Assuming that dividend yields proxy for growth expectations we find that investors are unduly optimistic about high growth stocks and too pessimistic about low expected growth stocks. These results support Lakonishok, Shleifer and Vishney's (1994) contention that contrarian investment strategies outperform the market because market participants have consistently overestimated future growth rates of glamour stocks relative to value stocks.