Existing studies show that firm asset and investment growth predict cross-sectional stock returns. Firms that shrink their assets or investments subsequently earn higher returns than firms that expand their assets or investments. I show that the superior returns of the low asset and investment growth portfolios are due to the omission of delisting returns in CRSP monthly stock return file and that the poor returns of the high asset and investment growth portfolios are largely driven by the subsample of firms that have issued large amounts of debt or equity in the previous year. Controlling for the effects of the delisting bias and external financing, I do not find an independent effect of asset or investment growth on stock returns.
We examine whether the firm-level asset growth effects documented in Cooper, Gulen, and Schill (2008) extend to the aggregate stock market. We find that aggregate asset growth is a strong negative predictor of future stock market returns. The return predictability is economically large and holds both in and out-of-sample. High aggregate asset growth is associated with more optimistic analyst forecasts and subsequent downward revisions, as well as greater earnings disappointments. In addition, aggregate asset growth provides complementary power to predict cross-sectional anomalies above and beyond the commonly used measures of investor sentiment. These findings suggest that the behavioral explanation for the firm-level asset growth effects extends to the aggregate level.