“A new study has concluded that mutual fund families often play favorites with their funds. They do this, it says, by giving certain funds extra shares of the most-sought-after initial public offerings and by arranging for funds in a family to take the opposite sides of certain trades, to the advantage of a chosen few of those funds,” wrties Mark Hulbert in Sunday’s New York Times.

“The study, called ‘Favoritism in Mutual Fund Families? Evidence on Strategic Cross-Fund Subsidization,’ was conducted by two finance professors at French business schools, José-Miguel Gaspar of Essec and Massimo Massa of Insead, with one of Professor Massa’s graduate students, Pedro Matos.”

“The research used a database of all actively managed domestic equity funds in the United States from January 1991 through June 2001 and focused on the 50 largest fund families, whose combined assets amount to more than 80 percent of assets in such funds. The study is circulating in academic circles as a working paper. The researchers found that, because of favoritism, some funds do better than they would otherwise, while others perform worse. The resulting performance gap averages 1.6 percent to 3.3 percent a year.”

“These conclusions are based on a statistical study of fund families as a group. Professor Massa says that the researchers have not tried to pinpoint favoritism at any fund family in particular, and that they do not know what might have motivated executives of fund families to engage in favoritism.”

“But they have suspicions. They say that because executives want to maximize total earnings, they favor funds that charge higher fees or funds that are within shouting distance of finishing at the top of year-to-date performance rankings. Many studies have shown that investors pour a disproportionate amount of money into funds that lead the rankings. One implication of this is that a fund family’s total earnings are likely to be higher if it has one highly ranked fund and one low-ranked fund than if it has two average performers.”

“The researchers found that executives of fund families used several approaches to favor certain funds. In one strategy, the highest-fee funds in a fund family, on average, received far more shares of initial public offerings than those with the lowest fees. And funds with the best year-to-date performances also received a disproportionately large share.”

“This pattern was even more pronounced for I.P.O.’s whose prices jumped the most on their first day of trading. These were presumably the I.P.O.’s for which market demand was greatest, and therefore could be most expected to increase the favored funds’ returns.”

“The researchers attributed an even bigger impact to a practice known as opposite trading, which occurs when funds within a family take the opposite sides of trades.”

“That can happen when the manager of a favored fund wishes to buy or sell an illiquid stock, Professor Massa said. The manager might otherwise think twice before making these trades, because the fund’s own buying or selling could hurt prices. But by having other funds in the family buy when favored funds are selling or sell when the favorites are buying, the manager of a favored fund can trade with minimal impact on prices.”

“Over the period studied, the stocks sold by favored funds generally lagged behind the market after they were sold, and the stocks they bought generally outperformed. As a result, opposite trading had the effect of transferring some of the returns of out-of-favor funds to their favored brethren.

Clearly, that was not in the interest of the shareholders in out-of-favor funds and, at the very least, raises questions about whether the managers and trustees of those funds have fulfilled their fiduciary responsibilities to shareholders.”

“While the researchers have no direct evidence of fund executives’ motivations, Professor Massa says the statistical evidence convinces him that it was not mere chance that a disproportionate share of I.P.O.’s during the study period were allocated to favored funds, or that the opposite trades between favored and out-of-favor funds regularly benefited the favored ones.”

“The researchers have not tried to determine the number of fund families that have played favorites. But Professor Massa said he suspects that more than just one or two fund families have done so.The researchers did find that, on average, favoritism is most prevalent at the 25 largest fund families. That can be explained by the greater opportunities for favoritism in bigger families, he said.”