Tweaks to the star rating system have made it better predictor of performance
Critics have long questioned the value of the Morningstarsystem, which awards five stars to top-performing funds and one star to those in the bottom of the barrel. In recent years, Morningstar has been tinkering with the system, and now a recent study by Morningstar indicates that the system does have some predictive value.
Under the Morningstar system, funds are rated according to their risk-adjusted returns. Funds that finish in the top 10 percent of their categories are awarded five stars.
Before 2002, Morningstar lumped funds into broad categories, such as domestic stocks and taxable fixed income. During periods whenfavored high-yield bonds and shunned government issues, nearly all the five-star ratings in fixed-income went to junk bond funds, despite their risks, while short-term government funds finished down in the standings. Investors who blindly bought five-star funds could be stuck with junk issues just before they collapsed. As a result, academic studies showed that the Morningstar system had no predictive value. Morningstar took heed, refining the process by narrowing the categories so that funds were only compared to their peers.
For its latest study, Morningstar selected funds that had five stars in June 2002 and examined their performance over the next three years. In general, five-star funds did better than funds with four stars or lower ratings. But the results were significantly better when Morningstar looked at five-star funds that had the lowest expense ratios. For example, five-star international stock funds with low expenses outdid 67 percent of their competitors during the three-year period. Domestic stock portfolios in the top quartile for low expenses outperformed 59 percent of peers.
While the star system may be helpful,head Russel Kinnel cautions that the stars should only serve as a first screen.