The rating should be used as a screen to narrow your list. Don't ask it to make your decisions. By Russel Kinnel, Contributing Editor March 31, 2007 One thing I've learned over my many years at Morningstar is that I can't write enough about the star ratings and fund-analyst picks. Each year in Morningstar FundInvestor I describe how the ratings and the picks performed and how best to use them. Yet the letters and e-mails keep coming: How do they work? What should I do with them?Star struck I understand why all the questions. The star ratings and picks are complicated and important, and they can be a huge help in building a good portfolio of funds. But you have to understand a little bit about how they work to make use of them. Here's the key: The fund-analyst picks are forward-looking, subjective recommendations of funds that our analysts think have the greatest chance of success. The star ratings are backward-looking, quantitative measures of past returns that are adjusted for risk and sales charges. Star ratings rank funds based on their relative performance within their category over the past three, five and ten years. We subtract sales fees from those returns, and we penalize funds that take a lot more risk than their peers (and reward those that take on much less risk). Our risk measure captures volatility with a particular emphasis on price declines. The result is a number that reflects each fund's risk and load-adjusted returns. We rank the funds by these numbers and award the top 10% five stars, the next 22.5% four stars, the middle 35% three stars, the next 22.5% two stars and the bottom 10% one star. In other words, we grade on a curve, and we don't rate everything a "buy" the way Wall Street does. Finally, we combine the star ratings for three, five and ten years into one overall rating. Advertisement We've made one big improvement. In 2002, we began comparing each fund only to its peers (funds in the same category) instead of to all U.S. stock funds or all international funds. We found that doing that gives the star ratings some predictive power, meaning that five-star funds as a group beat four-star funds, which in turn beat three-star funds, and so on down the line. The results so far indicate that five-star funds do better than average and one-star funds do worse, but only by modest amounts. The star rating's chief limitation is that it doesn't reflect recent fundamental changes, such as manager changes or asset bloat. The upshot of that and the performance studies is that the rating should be used as a screen to narrow your list. Don't ask it to make your investment decisions. Discriminating picks The fund-analyst picks start where the star ratings leave off. Whereas 10% of funds that have been around for at least three years receive five stars, only about 2% make our list of analyst picks. These are the funds that we believe have the strongest fundamentals and the greatest likelihood of success over the long haul. The most important fundamentals are management, strategy, long-term performance, costs and stewardship. We crunch numbers, interview fund managers and visit fund companies to figure out who has a real competitive advantage. Most of our picks rate four or five stars, but not all do. If a fund's fundamentals have drastically changed for the better or are just not fully captured by trailing returns, we'll make it a pick even if it has just one or two stars. Advertisement We've found that about two-thirds of our picks outperform their category over time. The complete picks list is available to FundInvestor subscribers and premium members of Morningstar.com. So the star rating is a good gauge of past performance that needs to be supplemented with fundamental research. The list of picks adds that in, enabling you to go right to the best in class. Columnist Russel Kinnel is director of mutual fund research for Morningstar and editor of its monthly FundInvestor newsletter.