NEW YORK (CNN/Money) - I'm looking for a good mutual fund for a long-term investment. How much weight do you think an investor should give to a fund's Morningstar star rating?

-- Vinny Joshi, Chicago, Illinois

In my opinion, not much. Oh, I know that the Morningstar star system, which assigns funds one (the lowest rating) to five stars (the highest) based on their past performance, is wildly popular among fund investors. So popular, in fact, that in many years four-and five-star funds attract more new money than all their lower rated siblings combined. But I believe that investors who choose their funds by gazing at the stars often end up star-crossed. To understand why I'm not a big fan of the star system, you've got to understand a bit about how it works.  

Morningstar begins by assigning each funds to one of four broad categories: U.S. and international stocks, taxable and municipal bonds. Each fund is then assigned a score based on its risk-adjusted return for the past three and, if available, five and ten years versus other funds in the same category. The funds are then ranked by their scores, with the highest scoring 10 percent receiving five stars, the next 22.5 percent four stars, the next 35 percent three, the next 22.5 percent two and the bottom 10 percent getting one lone star. The rankings are updated every month.

While this is all very precise, I believe the system tells you more about which types of stocks or bonds are doing well, as opposed to which managers are doing a good job running their funds. For example, when big growth stocks are all the rage, as they were in the late 1990s, large-company growth funds tend to hog the five-star ratings. And when small value stocks are in vogue, then small-company value funds garner a larger share of five-star rankings. In other words, managers whose performance may reflect nothing more than luck in being in a popular sector can achieve high ratings, while a manager who's an excellent stock picker but happens to be in a sector that's not in vogue can receive a lousy rating.

The other problem I have with Morningstar's star ratings is that they're not a very good barometer of future performance. One study, for example, showed that of the funds that had four- or five-star rankings in January, 1998, less than a third were able to maintain that ranking continuously over the next two years. In short, today's four- or five-star fund might be tomorrow's two- or three-star performer, although another study suggests that funds at the bottom of the rating scale tend to stay at the bottom.  

So, all in all, I don't recommend using the star ratings as a guide for choosing funds. I do think, however, it's worthwhile to compare a fund's performance versus that of funds in the same category -- that is, small-value funds vs. small-value funds, large-growth funds vs. large-growth funds, short-term Treasury bond funds vs. short-term Treasury bond funds, etc. This way, you get a sense of how a fund manager is faring compared with other managers pursuing the same investing strategy.

While you're at it, you should also take a look at the fund's risk level, using such barometer's as Morningstar's risk ratings. I also believe investors should factor expenses into their fund-investing decisions. While there will certainly always be high-expense funds that perform well, in general I believe that funds with average or below average expenses have a much better shot at generating superior long-term gains.

You can evaluate a fund using these criteria by getting a Morningstar Fund Report, available right from CNN/Money, or try using the Fund Screener tool. It takes a bit more work and thought choosing a fund this way, but I think you'll come up with better results than you will by simply looking at the stars.