News & Events

When Small Funds Aren't Necessarily Better Funds

When it comes to mutual funds, many investors think that small is beautiful. New research suggests that they are wrong.

By "small," I am not referring to funds that invest in stocks of smaller companies, the so-called small-cap stock funds, but instead to funds that are themselves small, having relatively few assets under management. Many people believe that such funds enjoy a performance advantage over the largest ones.

The belief stems in part from the relative nimbleness of smaller funds. It is difficult, for example, for a large fund to buy or sell a sizable position in some stocks without moving their prices up or down significantly. In some cases, it would be downright impossible for a big fund to establish a position that it would consider worthwhile.

Imagine yourself in the shoes of Robert Stansky, manager of the Fidelity Magellan fund, which has $78 billion under management. His largest current holding is General Electric (news/quote), of which Magellan owns $3.7 billion in shares. That amount exceeds the market capitalization of even the largest company in the Russell 2000 index, a benchmark for smaller stocks, and it is greater than the market caps of about half the companies in the Russell 1000, a large-cap benchmark.

So when it comes to around 2,500 of the 3,000 companies in these two Russell indexes, there is no way Mr. Stansky could buy as large a position as his stake in G.E., no matter how much he likes it. (By the way, these 3,000 companies represent more than 98 percent of the United States equity market.)

Even among many of the 500 remaining companies, Mr. Stansky would still have difficulty buying a position as large as his G.E. holding. The average daily trading volume of many of these 500 companies is but a small fraction of $4 billion, so any attempt to add such a large position would so distort the market as to make the trade untenable.

Mr. Stansky could still establish smaller positions in many of these companies. But he would face a trade-off: the smaller the stake in a given company, the smaller its potential contribution to his fund's profitability.

Small mutual funds, in contrast, are at a significant competitive advantage. Consider a fund that has $25 million under management -- which would hardly be the smallest fund in the universe. That level of assets, however, is just 0.03 percent of Magellan's assets. Virtually no publicly traded stock would be beyond the reach of a fund that small.

The ability to trade in small-cap stocks lets the smallest funds exploit the so-called small-cap effect -- the historical tendency of small-cap stocks to outperform large ones. Since 1926, on average, that performance difference has been about 1.2 percentage points a year, according to Ibbotson Associates. To qualify as a small-cap stock, under the definition of Ibbotson and others, a company must have a market capitalization of less than $200 million.

For years, the data confirmed the theory that small funds perform better than large ones. One study, for example, found that for every doubling in a fund's size, investors could expect a decrease of 21 basis points, or hundredths of a percentage point, in annualized performance. That adds up to a big difference, especially when considering that the largest funds are many times bigger than the smallest.

This is where the conventional wisdom stood until the late 1990's, when a fatal flaw was found in the historical comparisons of large and small funds. The discovery was made by Mark M. Carhart, then an assistant professor of finance and business economics at the University of Southern California and now co-head of quantitative research at Goldman Sachs (news/quote) Asset Management. Mr. Carhart found that the previous studies had used historical databases that ignored funds that had gone out of business.

That omission was significant, and for two reasons: Smaller funds go out of business much more often than larger ones, and funds that fold tend to be poorer performers than those that survive. Upon correcting this so-called survivorship bias, Mr. Carhart could detect no systematic difference in the performances of large and small funds. (His research has circulated for several years as a working paper: http://papers.ssrn.com

/sol3/papers.cfm?abstract-id=36091).

To understand his findings, consider that the average small fund is riskier than the average large one -- so risky, in fact, that there is a much greater chance that it will lose enough money to force it out of business. After all, when you bet aggressively, you have higher odds of winning big -- and losing big. On average, you end up no better off than you would have by going with a larger fund.

So where did conventional wisdom go wrong? Why can't the average small fund capitalize on its abilities to trade with lower transaction costs and to invest in smaller-cap stocks?

One major reason is now known: The average smaller fund spends a greater percentage of assets on administrative costs, advisers and managers, and on marketing costs like 12b-1 fees. In fact, according to research conducted several years after Mr. Carhart's paper, the smallest 25 percent of funds have average expense ratios that are 94 basis points higher than those of the largest 25 percent of funds. (This research was conducted by John M. R. Chalmers of the University of Oregon, Roger M. Edelen of the Wharton School of the University of Pennsylvania and Gregory B. Kadlec of Virginia Tech.)

Even with a higher expense ratio, a small fund still has fewer dollars to spend for luring a top money manager and buying the best research. To be sure, a superstar manager and expensive research do not guarantee success. But Mr. Carhart's finding suggests that they generally help.

Yet another factor diminishes the theoretical advantage of the smallest funds. The small-cap effect, it seems, is not as exploitable in practice as theory suggests, even by the smallest funds. Recent research by three finance professors -- Joel L. Horowitz of Northwestern, Tim Loughran of Notre Dame and N. E. Savin of the University of Iowa -- has found that the small-cap effect is being driven almost completely by the tiniest companies.

How tiny? The three professors found that if companies with market capitalizations of $5 million or less are eliminated from the record, the small-cap effect disappears. That is devastating to the small-cap effect because it is virtually impossible to make money by investing in such companies -- they hardly ever trade, and their bid-ask spreads are prohibitive.

The bottom line is this: In choosing a mutual fund, you should not take its own size into account. Instead, search for a fund that has beaten the market over many years. If it happens to be a small fund, so be it.

By MARK HULBERT


Return to top of page