Falkenstein documents the apparent preferences of U.S. open-end mutual fund managers for various stock characteristics. Using Morningstar's data on mutual fund portfolio holdings for the years 1991 and 1992, the author calculates the relative holding of shares by the mutual fund sector as a whole and by specific fund objective. Although the sample has only two years of data (and more than 1,000 mutual funds for each year), it is free of survivorship bias because it was not constructed retroactively by Morningstar.

The author's results show that mutual fund managers display a nonlinear preference for stocks with high volatility. This finding is consistent for all subsectors of mutual funds and stocks studied. Relevant to transaction costs, mutual fund managers show an aversion to low-priced stocks and a preference for stocks with high liquidity. Low-priced stocks are known to have high bid-ask spreads as a percentage of their average price. The greater transaction costs of low-priced stocks provide a reason for mutual fund managers to avoid them.

Other than small-capitalization funds, which specialize in small firms, most mutual fund managers show an aversion to small firms. Several possible explanations exist for this phenomenon. Many mutual funds that declare themselves to be “diversified” are restricted from owning more than 10 percent of a firm's voting shares. This restriction puts a constraint on a fund manager's demand for small-cap stocks. Furthermore, owning a large proportion of a firm's outstanding stock implies that a premium must be paid to put on or take off a position. Therefore, large funds may be discouraged from buying stocks of small firms because they perceive that significant transaction costs are associated with buying and selling these shares compared with large-cap stocks.

Mutual fund managers also tend to avoid stocks with low levels of information, as measured by the number of major newspaper articles on the firm or the number of months since the stock's listing on the exchange. These low-profile stocks entail high search costs to highlight them as securities desirable for a portfolio. Estimating risk for these stocks may also be difficult and inaccurate.

The author also addresses tests of herd behavior and trend following by mutual fund managers. Herding refers to any mass movement toward particular stocks. Trend following is a specific type of herding that involves a large group of fund managers chasing stocks that have recently risen in value. If mutual fund managers prefer to hold securities with certain characteristics, then as a security acquires these characteristics, mutual fund managers in aggregate will purchase it. Thus, the author's empirical findings provide an alternative way for researchers to look at herding and trend following and suggest variables to control for when conducting such research.