Don’t Navigate Only by the Stars, Study Warns

Morningstar ratings have become the Good Housekeeping seal of approval for mutual funds. Investors frequently scan a fund’s rating before sending in money, and investment companies tout portfolios with the highest marks in their ads and shareholder announcements.

But you should take these ratings with a grain of salt, says an Ivy League finance professor in a new study. The reason: Not all funds have an equal shot at earning a top mark.

Marshall Blume, a finance professor at the University of Pennsylvania’s Wharton School, says the rating system developed by Morningstar Inc. of Chicago discriminates against certain types of funds. In a study of the ratings of nearly 2,000 domestic stock funds, he reached the following conclusions:

* Funds with lengthy track records have less of a chance of receiving a top, five-star grade than newer funds.

* Sector funds, which invest in specific industries, are less likely to score a top grade than their more diversified cousins.

* Broker-sold funds stand less of a chance of earning a superior mark compared with funds that carry no sales charges.

Of these assertions, the last is probably the most controversial. Morningstar penalizes load funds by subtracting the size of their sales charges from performance, giving an edge to direct-marketed funds.

“Load funds are penalized, and it’s hard to say whether that’s good or bad,” Blume said in an interview. “It depends on what you get as a benefit, in terms of advice and service from a broker.”

Paul Pudaite, Morningstar’s director of quantitative research, defends the company’s lumping together of load and no-load funds and penalizing the latter. “We want to report figures that are relevant to investors,” he said. “Those loads will be charged, and shareholders will pay them.”

Barbara Levin, executive director of the Forum for Investment Advice, a Bethesda, Md., group that represents load-fund companies, says it’s a tricky issue.

“Technically and numerically, what Morningstar does is fair,” she said. “But by factoring in the sales charge, there’s an assumption that it’s just another fee tacked on, without recognizing the additional help rendered.”

In other words, if you pay a financial planner to choose no-load funds for you, those fees are not factored into the performance of a mutual fund. But a sales charge is factored in by Morningstar, even though it is often an alternative way to pay for such advice.

Levin added that people who invest hundreds of thousands of dollars will avoid some, if not all, of the sales charge, meaning that their returns could be higher than those implied by a Morningstar rating.

One of Blume’s major gripes is that Morningstar compares too many types of domestic stock funds against one another, including sector funds, convertible bond funds and stock-bond “hybrid” funds.

Since sector funds tend to be fairly volatile--and because Morningstar’s ratings penalize high volatility--these industry-focused portfolios fare poorly, he said.

“Morningstar tends to assign low ratings to these funds, which increases the number of stars assigned to other funds,” Blume said.

A better system would grade funds only against similar types of funds, he said. “You wouldn’t want to compare health against technology funds, or small-value funds against large-growth portfolios.”

Steve Lipper, a vice president at Lipper Analytical Services in Summit, N.J., noted that Morningstar also lumps together a range of bond funds, including such dissimilar products as short-term Treasury and junk-bond funds.

“Our basic point is that performance evaluation is best done using comparable categories and that the wider the category, the less reliable a rating is,” he said.

Lipper Analytical ranks funds on a performance basis only.

Pudaite acknowledged Blume’s criticism as valid but countered that Morningstar’s approach helps keep investors from straying too heavily into sector funds, which often top the short-term performance charts yet can be quite risky.

Blume’s study observes that no-load funds holding a broad mix of U.S. stocks come out smelling like a rose.

Nearly half of these funds in the survey attracted either a four- or five-star grade, but only one-quarter received a one- or two-star mark. Based on Morningstar’s bell-curve rating system, you’d expect that roughly the same number of funds would receive superior ratings as would get inferior grades.

Blume also questioned Morningstar’s basis for judging a fund’s riskiness, because it measures only downside volatility. A better risk gauge, he said, is a measure known as standard deviation, which tracks both upside and downside fluctuations. Since investors are more interested in the downside risk, however, there is some logic to Morningstar’s approach.

Blume’s assertion that older funds have more trouble earning five-star ratings reflects the fact that Morningstar mixes performance records for different time periods.

Assuming a fund has been around at least 10 years, its rating will reflect a combination of its three-, five- and 10-year performance numbers. Conversely, a 3-year-old portfolio would be judged only by its three-year record.

This blending effect for older funds tends to reduce the number of extreme scores, Blume says. That is, older funds generate not only fewer five-star ratings but fewer one-star grades as well. But since five-star funds receive more attention than one-star funds--especially in ads and company announcements--this tends to harm older portfolios on balance, Blume argued.

In his view, Morningstar’s system is a messy one that makes direct comparisons difficult.

“I don’t think you should publish three-, five- and 10-year numbers in one rating,” Blume said. “It’s like mixing apples and oranges.”

Although Blume’s study reveals shortcomings in Morningstar’s approach, it’s safe to say that it’s likely any rating system would have some imperfections. Indeed, some of his criticisms also apply to the ratings of the rival Value Line Mutual Fund Survey, which Blume did not examine.

Still, the study serves as a reminder for investors not to rely too heavily on a rating for buy or sell signals. Mutual funds are simply too complex to be judged by a single number.