Stock Price Volatility Tests of Narrow and Broad Dividends: A Closer Examination of the Results of Ackert and Smith

Abstract

Ackert and Smith (1993) suggest that volatility tests of stock prices should be based on a more inclusive definition of dividends than has actually been used in past studies. They perform a West Test using such a dividend definition, and conclude that stock prices are not excessively volatile relative to these dividends. This paper considers their work on two bases: First, it argues that in most cases the expanded dividend definition would not be expected to lead to volatility conclusions that differ from those reached under the narrower definition; second, it presents evidence that Ackert and Smith's empirical results may have been induced by unrecognized test bias created by non-stationarities in their data.