Avoidance of Small Stocks and Institutional Performance

Recent empirical literature finds that institutions as a group outperform the rest of the market, before costs. This paper argues that institutional strategy in small stocks during 1980s and 1990s was largely driven by factors other than expectations of stock returns. We propose a decomposition of institutional excess return into three components: stock picking within a particular class of stocks, capital allocation across stock classes, and class timing. The capital allocation component is dominated by small stocks - institutions consistently avoided small stocks, and small stocks consistently underperformed the rest of the market, which contributes around 30 basis points to the total institutional excess return of 100 basis points per year. However, the underperformance of small stocks is dominated by stocks with low return on equity (ROE) and new stocks for which accounting data is not available yet. We show that institutions consistently avoided all groups of small stocks without much regard for ROE: For the sample period 1982 - 1997, only 3 basis points of institutional excess return come from underweighting of low-ROE and new stocks, while the remaining 25 basis points come from allocating capital away from small stocks without any regard for ROE. Had institutions avoided small low-ROE and new stocks while still allocating the same share of their capital into the two smallest size quintiles, they would have increased their excess return over the rest of the market by 17 basis points per year, which is a significant part of 100 basis points of historical excess return. This provides suggestive evidence that institutions avoided small stocks for reasons that were not related to their expectations of small stock returns.