On the Scope and Drivers of the Asset Growth Effect

Abstract

Recent papers have debated whether the negative correlation between measures of firm asset growth and subsequent returns is of little importance since it applies only to small firms, justified as compensation for risk, or evidence of mispricing. We show that the asset growth effect is pervasive and evidence to the contrary arises due to specification choices; that one measure of asset growth, the change in total assets, largely subsumes the explanatory power of other measures; that the ability of asset growth to explain either the cross section of returns or the time series of factor loadings is linked to firm idiosyncratic volatility; that the return effect is concentrated around earnings announcements; and that analyst forecasts are systematically higher than realized earnings for faster growing firms. In general, there appears to be no asset growth effect in firms with low idiosyncratic volatility. Our findings are consistent with a mispricing-based explanation for the asset growth effect in which arbitrage costs allow the effect to persist.