Earnings Dispersion and the Stock Market Reaction to Aggregate Earnings News

Abstract

Using monthly and quarterly cross-sectional dispersion in firm level earnings news as a proxy for investor uncertainty about the implications of current aggregate earnings for future discount rates, I find that higher investor uncertainty leads to a lower stock market reaction to aggregate earnings surprises. An increase in the monthly cross-sectional dispersion in seasonally differenced earnings from the 25th to the 75th percentile of its time series distribution renders the contemporaneous market return−aggregate earnings relation positive and even significant depending on the measure of earning used. The stock market reaction to aggregate earnings surprises becomes insignificant when earnings data for the 2008-2009 period are included in the analysis. This suggests that the most recent economic crisis has distorted the aggregate earnings-market return relation which has generally been found to be negative in the prior literature. Finally, in contrast to Jorgensen, Li, and Sadka (2009) and consistent with the predictions of the modern portfolio theory, I find that there is no relation between dispersion in earnings surprises and stock market returns using monthly and quarterly time series of seasonally differenced earnings and dispersion in these differences. This relation depends on the magnitude of the earnings change: significantly negative (positive) if the earnings change is in the bottom (top) quartile of its time series distribution.