Morningstar’s stars shine after overhaul

For years, critics have slammed the Morningstar rating system for elevating faddish offerings while ignoring out-of-favor funds toiling away in difficult environments. But since Morningstar Inc., a Chicago investment research firm, overhauled its widely watched star ratings, at least two academics believe that the stars now actually may be worth something. Strong performance Matthew R. Morey and Aron Gottesman, associate professors of economics at New York’s Pace University, recently published a study in which they found that the higher a fund’s star rating, the better its performance three years out. They concluded that the stars actually are meaningful in predicting future performance of a fund — at least for a while. Prior to the current Morningstar rating system, which went into effect in mid-2002, the stars “were terrible predictors of performance,” said Mr. Morey, author of an earlier study that found no difference in future performance between a five-star fund and an average fund. The rating system overhaul, he said, has gone a long way toward making the star ratings meaningful. “At least for the three-year period, the star ratings did predict a lot better in terms of risk-adjusted performance,” Mr. Morey said. “There is some evidence that performance does persist, but whether it’s one year or two years, we’re not sure.” In his analysis, Mr. Morey found that five-star funds are likely to outperform four-star funds, which are likely to do better than three-star offerings and so on. He attributes the results largely to the new, narrower categories in which funds now find themselves. The old methodology compared funds in just four broad asset classes — domestic equity, international equity, taxable bond and tax-exempt bond. There now are 48 categories, so funds are much more likely to be judged against similar funds. “You had a situation in 1999 where the worst growth fund had a higher rating than the best value fund,” Mr. Morey said. “So all a growth manager had to do to be a four- or five-star fund was beat a value manager.”
Today, a growth fund sports a five-star rating only if it outperforms other growth funds on a risk-adjusted basis, rather than by being in a hot market segment. The same is true for value offerings. These are the kinds of results that Morningstar had hoped to achieve when it undertook the change, said Don Phillips, a managing director at the firm. “The real utility of the star ratings is shining through,” he said. Although advisers applaud the change, many still believe that the star ratings are simplistic and confuse clients, who tend to equate more stars with more-talented managers. “Only about 25% or fewer or our clients pay any attention to the ratings, but those who do place a high value on them,” said John Brown, founder and principal of Brown Financial Advisory Inc., an advisory firm in Fairhope, Ala., that manages $100 million. He finds it a struggle to get clients to understand what the stars actually connote. “Back in the late ’90s, [real estate investment trust] mutual funds had two stars or one star, and we were recommending them because of the diversification they bring,” Mr. Brown said. “It was hard explaining it.” A starting point Mr. Phillips was quick to point out that “the last thing in the world we want to do is tell people to blindly follow the star ratings. But it’s a rational starting point. With 10,000 choices out there, you need some kind of screening device.” Michael Kresh, president and chief investment officer of M.D. Kresh Financial Services Inc. in Islandia, N.Y., looks at star ratings in a different way. “I shy away from five-star funds,” he said. “I’d much rather put a client in a three-star fund that’s on the way to being a five-star fund than a five-star fund on its way down.” Funds with average ratings, Mr. Kresh said, fly below the radar and have relatively low inflows. They may be solid contenders that can continue churning out respectable long-term performance without the distractions of being in the spotlight, he said.