Morningstar's famous fund rating system turns out to have been surprisingly predictive, according to one study. Good to know, but the future isn't necessary written in those stars. By Thomas M. Anderson, Contributing Editor September 14, 2006 Thanks in large measure to Morningstar, a Chicago financial-information company, many investors treat mutual funds like restaurants. This fund has four stars, so it must be decent. That fund has only one star, so you want to pass on it. Advertisements from fund companies and the herdish instincts of the media encourage this Zagat-like thinking. But what can Morningstar's catchy and convenient ratings really tell us? A lot more than they used to, says Matthew Morey, a finance professor at Pace University, in New York City. He and his colleague Aron Gottesman examined all the U.S. stock funds rated by Morningstar as of June 2002. That's when Morningstar changed the recipe of its ratings to what they are today. Morey and Gottesman tracked performance of these funds for three years. They found that in the improved system, higher-rated funds outperformed lower-rated funds for the most part. Five-star funds did better than four-star funds, four-star funds outpaced three-star funds and so on. "I didn't expect the results we got," says Morey, who will publish his study in a forthcoming issue of the Journal of Investment Consulting. "The previous rating systems had flaws in its methodology." Morningstar's ratings are based entirely on the numbers. The firm assigns stars based on a fund's past risk-adjusted results, taking into account any sales charges a fund levies in calculating performance. Funds must have a record of at least three years before qualifying for a star ranking. Those at the zenith of the rankings receive five stars; those at the bottom get one. In 2002, Morningstar revamped its rankings in two major areas. First, and most important, it switched from four ranking groups to 48 different categories. Instead of lumping all U.S. stock funds together, they are sliced and diced them into categories such as small-cap value and large-cap blend. Under the old regime, funds that focused on the hottest sectors of the market usually received the most stars. For instance, in mid 2000, three-quarters of all funds that invested mostly in large, fast-growing companies garnered four or five stars. At the same time, only six of the 123 funds that specialized in small, bargain-priced stocks earned four or five stars. Since then, large-company growth funds have tanked, while small-company value funds have prospered. Now stars are dolled out more evenly among the categories. For example, a large-company growth fund can earn five stars even if its performance relative to the overall market as well as to funds in other categories is unimpressive. Secondly, Morningstar recalibrated how it judged the risk of a fund. It used to measure risk as the fund's average performance relative to the return of the riskless 90-day Treasury bill. If the fund beat the T-bill benchmark, it was deemed safe. This means funds with highly variable returns could be called low-risk by Morningstar as long as they returned more than a T-bill. High-flying, highly volatile technology funds proved the folly of this approach. In 1999, these funds displayed little risk by Morningstar's standard, Morey says. But during the 2000-02 bear market, tech funds suffered huge losses. Now Morningstar measures risk by assessing the month-to-month variations in a fund's returns. Despite the tweaks, the Morningstar system still has notable deficiencies. It doesn't consider manager changes, and it ranks older funds over longer periods of time than it does newer funds, which can create an apples-and-oranges comparison. The system in effect takes expenses into account because ongoing fees are reflected in total returns. But because no fund can assure that previous performance will be repeated in the future, it would be better if the rankings penalized high-expense funds and rewarded low-cost funds. Fortunately, Morningstar frequently warns investors to use its rankings as the beginning of a fund search, not the end. "I think people are doing a better job with the star ratings," says Russel Kinnel, Morningstar's director of mutual fund research. "We often discuss the rating's predictive value and its limitations." Morningstar goes beyond the star ratings by having its analysts pick their favorite funds. These choices reflect how analysts think the funds will perform in the future. Morningstar's 172 top picks include 13 two-star funds and two one-star funds. Analyst picks are only available to viewers with a Morningstar subscription. Lazy investors who rely solely on Morningstar ratings may be relieved by the Pace University research. But Morey and Gottesman caution that the predictive power of the star system could diminish over time. "We only have three years worth of data," Morey says. Want to see our favorite fund picks? Check out the Kiplinger 25.