By rushing in, people change a fund in ways that may inhibit its ability to repeat the past. By Russel Kinnel, Contributing Editor May 1, 2007 In the fund world, success is a dangerous thing. Strong returns grab investors' attention and bring a torrent of new cash for the manager to invest. The benefits of that extra money may include falling operating expenses. But as more and more cash floods in, a hidden expense -- the cost of trading -- begins to rise. In addition, the manager may be forced to change his or her investing style to adapt to the fund's new girth. That's the irony. People rush in, hoping for a repeat of the past. Yet by rushing in, people change a fund in ways that may inhibit its ability to repeat the past. I'll cite four examples.Burdens of size When a fund gets too big, it's rarely a disaster -- typically, it spreads its bets out over more stocks, and performance becomes more like that of an index or the typical fund in the category. I can think of much worse results than that, but who picks a fund with the idea of enjoying average performance? Plus, asset bloat doesn't just lead to greater diversification. It can increase trading costs because the fund's trades push a stock's price further away from the initial trading price. It can also lead the manager to alter strategy by investing in more-liquid stocks, trading less and, as mentioned, diversifying more -- all virtuous things, but probably not what made the fund excel. Royce Total Return is one of the largest small-company funds in existence, yet it remains open. From 2000 to today, the fund has grown from $280 million in assets and 150 holdings to $6 billion and 450 holdings. In addition, lead manager Chuck Royce has about $5 billion more under management in other funds. On the plus side, Royce uses a low-turnover strategy, which is a little less sensitive to asset bloat. Even so, this fund will be hard-pressed to match its past strong results because any single idea will contribute only a small amount to the fund's returns. Indeed, the fund's three-year returns are already subpar. Calamos Growth got too big, too fast, and that led it to underperform. It's working on its third straight below-average year. Although the fund isn't a pure mid-cap vehicle, investing even half of its $16.5 billion in mid caps is a challenge. Most of the fund's great years came before it crossed the $1-billion line, and its poor performance has spurred redemptions. If Calamos has learned its lesson, it will close the fund when cash flow again turns positive. Advertisement With a mammoth $167 billion in assets, American Funds' Growth Fund of America is in uncharted waters. GFA is better equipped to handle asset bloat than, say, former superfund Fidelity Magellan because its ten named managers operate independently, and it farms out a portion of the fund to individual analysts. American boasts the deepest team of managers and analysts around, yet its supply of experienced professionals and good, liquid stock ideas is not infinite. Although closing to new investors would only slow the inflow of money, it would be prudent to err on the side of protecting existing shareholders and putting off the day when the quality of management gets seriously diluted by the weight of assets. Up the size staircase At $19 billion, Fidelity Value is nearly twice the size of the next-largest fund in the mid-cap value category and four times the size of the next-largest fund in that category that's still open to new investors. Manager Rich Fentin didn't increase the number of holdings to accommodate asset bloat. Instead, he bought larger companies, so the fund now invests about half its assets in large caps. Fidelity Value never claimed to be a pure mid-cap play, but it has departed from the territory that drove past returns. In addition, Fentin is unable to add to some of his favorite names because the fund company is maxed out on the amount (15%) of a single stock it can own. Columnist Russel Kinnel is director of mutual fund research for Morningstar and editor of its monthly FundInvestor newsletter.