One-Year Rating System May Lead to Star-Crossed Investments

When it comes to investing, “What have you done for me lately?” is a dangerous question.

It leads to short-term thinking, which leads to searching for patterns where none exist--and that leads to bad decisions.

Think about it in terms of what psychologists call the “gambler’s fallacy,” in which you start seeing things that aren’t there.

Say you have flipped a coin four times, and each it has come up heads; the gambler believes that either heads is on a hot streak or tails is overdue. Somehow, that is supposed to make one or the other more likely to come up on the fifth toss.

But we all know--or should know--that there will always be a 50-50 chance of what happens next; the odds never change, no matter the recent pattern.

Short-term investment decisions and misuse of past performance data were the main themes at the Second Congress on the Psychology of Investing held in Boston last month. The issues were relevant for mutual fund investors because earlier this month Morningstar Inc. reluctantly unveiled a short-term performance indicator that is as likely to help investors as the flip of a coin.

The measure is a one-year star rating, a version of Morningstar’s familiar star system that is based entirely on what a fund has done in the last year.

Morningstar is an independent research firm that rates funds on risk-adjusted performance, meaning how well a fund does based on the types of risks it takes. The Chicago-based firm has become the major source of fund information for retail investors; more than 95% of all monies flowing into funds goes to those carrying Morningstar’s four- and five-star ratings.

The star system is part of a Morningstar fund review that includes a wide range of key data. Much as Morningstar itself has downplayed the star system, the money flows are evidence that stars play a big role in how people pick funds.

Until June, Morningstar gave its ratings based only on longer-term performance, no fewer than three years. So, why did it add the one-year rating?

About two years ago, the National Assn. of Securities Dealers changed the rules governing how funds advertise, prohibiting the use of selective time periods where a fund would show itself only during times when it looked good. The NASD also sought consistency, so that funds would display the same numbers to create an apples-to-apples comparison.

Under the guidelines, fund advertising should include one-, five- and 10-year performance numbers.

At first glance, Morningstar did not seem to be affected by the guidelines, since it is a research firm and not a mutual fund.

But financial planners make up about 80% of Morningstar’s client base, and those planners distribute Morningstar fund evaluations to customers. And, until this month, those evaluations had just five- and 10-year stars, not the one-year number.

Since financial advisors show Morningstar data to prospective fund buyers, the NASD feels that the company’s products must be treated like fund advertising. As such, Morningstar either needed to add the one-year rating or advisors had to stop using the sheets--the former a stupid choice, and the latter a suicidal business decision.

If it seems as if the regulators are overstepping their bounds, it should.

Morningstar issues independent research opinions. Under the same rule, any newsletter that looks at “bull market” and “bear market” fund performance would also be out of bounds, assuming that a sales rep reproduced the numbers and showed them to a potential investor. In fact, certain things that you find in a column like this one might violate the rules if it was redistributed by a financial advisor.

There comes a point where the extremes look pretty silly--and the NASD-Morningstar situation is there now.

Negotiations between Morningstar and the NASD, by most accounts, did not go well. But Morningstar President Don Phillips acknowledges that he thought the quickest way to make the whole thing go away was to “throw in the towel and do this and then have everyone realize that this is the dumbest idea ever.” In other words, he made the decision to give the NASD what it wanted, and now he’s stewing in his juices while the industry decides what, if anything, to do next.

Thus far, wiser heads have not prevailed. Even if the fund industry pressures the NASD to change the rules, that would require Securities and Exchange Commission approval, a process that moves with all the speed of a glacier.

Morningstar and the NASD will continue to slug it out, each offering compromises that the other is likely to reject, at least for the time being.

It would be a shame for Morningstar to lose its reputation--and for investors to lose one of their most trusted tools--because it has to send out data that really can’t be trusted.

In the meantime, one-year star ratings could mess up investors. Here is why they are a poor indicator:

* Say you have a mutual fund that invests in something risky, like gold. The fund goes up by at least a fraction each month. Statistically speaking, that fund is risk-free, which means that its risk-adjusted return--the number Morningstar uses to base its ratings on--might be skewed.

* The one-year ratings penalize load funds by taking out the entire load. Longer-term ratings allow the fund the time needed to recoup the sales charge and put up good numbers, but the load is a tough nut to overcome in just one year, greatly reducing the chance that a load fund will get a good rating.

* The short time period is sure to put some lousy funds on the top of the heap, at least for a little while.

* Given the general public’s reliance on star ratings, the one-year numbers could sway a lot of decisions.

At the investment psychology conference, Mark Hulbert of the Hulbert Financial Digest noted that his studies have shown that the correlation between one-year rankings and future performance “is zero. One-year numbers have no real statistical significance.”

The answer for investors is to ignore the short-term star rankings. Morningstar is distributing the one-year ratings on materials it sends only to financial advisors, but don’t let them convince you that this is somehow “new and improved” data, when it is actually a poorly conceived shell of a respected evaluation system.

What a fund has done lately is not what counts; what a fund will do tomorrow is what you’re worried about--and you can’t divine that by looking at the stars.

Charles A. Jaffe is columnist at the Boston Globe. He can be reached by e-mail at jaffe@globe.com or at the Boston Globe, Box 2378, Boston, MA 02107-2378.