In the fund world, though, it pays to rally behind losers.

Here�s how it works: At the end of every year, we find the three least popular fund categories of the year based on percentage change in cash flows. We then recommend that you buy one fund from each unpopular category and stick with them for three years.

Morningstar has backtested the strategy to 1987, with winning results. Unpopular categories beat the average equity fund over the next three years more than three fourths of the time.

The odds versus popular categories have been even better. Unpopular categories have topped the popular ones over the next three years more than 80% of the time. Investing doesn�t get much closer to a sure thing than that.

This year, our strategy means buying from the Latin America, precious-metals, and convertibles categories.

The Rules

Putting our unpopular-categories strategy into action is simple. In fact, there are just three rules to follow:

Buy one fund from each category. Staking everything on just one unpopular category can be risky. Not every unloved category will catch fire, and one category can pull the weight for the entire group. The unpopular categories from 1988 averaged a 19.2% annualized return for the three subsequent years, for instance, but the health-care category�s 37.7% annualized return was crucial to that performance.

Buy before summer vacation. Readers have asked whether they have to buy right away for the strategy to work. Good news: After examining category returns for the three-year holding period minus one month, minus two months, and so on, we found that there�s not much of a penalty for lateness. Most of the time, you could buy the unpopular categories as much as a year after reading our story and the strategy still worked. We�d rather err on the side of caution, though. Buying the funds at the beginning of June worked as well as buying them in January, so just be sure to purchase by midyear.

Limit your bets. Resist the urge to put more than 5%--or at most 10%--of your portfolio into unpopular categories. That way you�ll minimize the disappointment in one of those rare occasions when the strategy doesn�t work. We advocated holding precious-metals, communications, and Europe-stock funds from 1996 through 1998, for example. Precious-metals funds lost an annualized 18% for those three years, dragging the unpopular average down to 10%. The popular technology, financials, and mid-cap growth categories, meanwhile, rumbled on to an average annual return of more than 20%.

Three Ways to Use the Strategy

Invest new money. Maybe your company paid you a nice bonus this year, or you have some cash you�ve been sitting on in case of Y2K troubles. Put that money to work by buying one fund from each of the Latin America, precious-metals, and convertibles categories.

Sell one, buy the other. If the holiday season left you without much excess cash, try cutting back on popular fund categories and shifting the gains into out-of-favor categories.

This year, that means scaling back on Japan, technology, and Asia/Pacific ex-Japan funds. There are two exceptions here: If you�ve been following our strategy, you bought an Asia/Pacific ex-Japan fund last year. Don�t sell it because it was popular in 1999. Hold it for the full three years, but consider taking profits. (Hold it long enough to avoid short-term capital gains, though.)

Alternatively, one year�s popular category could land among the next year�s unpopular clique. In 1992, for example, investors loved financials. In 1993, investors turned their backs on the category, making it the second least popular. Rather than sell a category one year only to turn around and buy it back the next, don�t sell your entire stake in popular categories. Instead, just take profits. So if next year we recommend buying Japan, technology, and Asia/Pacific ex-Japan funds, you can just add new dollars to your existing holdings.

Cut back on the favorites. You�ll do yourself a favor by simply reducing your exposure to popular categories. Had you followed our advice in early 1998, you would have avoided real-estate and small-value funds� losses of 16% and 7%, respectively, and financials� subpar 6% gain in 1998. None of those categories did better than mark time in 1999.

Latin America

Scudder Latin America SLAFX The managers like companies that dominate their markets, generate lots of cash, and have little debt. They believe such firms have a better chance of weathering Latin America�s frequent economic storms. Not many stocks meet those criteria, so the managers spend a lot of time trading in and out of them--buying on dips and selling on rallies.

Van Kampen Latin America MSLAX This fund invests in Latin American blue chips, emphasizing countries with favorable macroeconomic trends. Like many Latin America fund managers, this team trades frequently to take advantage of the often-dramatic fluctuations in stock prices. Buy the C shares; they�re the best deal for a three-year time horizon.

Fidelity Latin America FLATX After a tough 1994, manager Patti Satterthwaite overhauled the fund�s strategy. Broad country bets were out, stock-picking was in. Satterthwaite now looks for well-established companies that are producing a lot of cash and don�t carry much debt. So far, the makeover seems to have suited the fund well. This fund carries a 3% front-end load.

Precious Metals

American Century Global Gold BGEIX This straight-ahead offering is essentially an enhanced index fund. Because it uses the Financial Times Global Gold index as its benchmark, the fund focuses on bigger companies and diversifies among more countries than do other precious-metals funds. This fund keeps costs low--it�s even cheaper than Vanguard�s precious-metals offering.

Vanguard Gold & Precious Metals VGPMX One of the best precious-metals fund around. The key is manager Graham French�s conservative approach to this volatile sector. He emphasizes established producers, and includes gold and platinum bullion in his portfolio. French�s careful style has also steered the fund clear of blowups, such as 1997�s Bre-X scandal. And then there are Vanguard�s habitually low expenses.

Oppenheimer Gold & Special Minerals OPGSX Manager Shanquan Li is another conservative gold investor, tending to focus on low-cost producers with extensive reserves. The fund gets additional stability from a 20% stake in nongold firms, such as platinum and palladium producers. Li�s caution has made this one of the smoothest rides in a very volatile category. Go with the C shares.

Convertibles

Davis Convertible Securities RPFCX This fund shares the value bias of its equity siblings. While other convertibles fund managers have loaded up on technology, manager Andrew Davis has dedicated a lot of this portfolio to REIT and energy convertibles. The fund, which carries a 4.75% load, nonetheless has competitive long-term returns.

Oppenheimer Convertible Securities RCVAX Co-managers Michael Rosen and Edward Everett invest temperately, aiming for a portfolio that�s less sensitive to the underlying stocks. The portfolio�s credit quality is higher than the average, and includes companies with strong balance sheets and improving cash flows. The C shares work best for a three-year time horizon.

Fidelity Convertible Securities FCVSX This fund has a strong track record despite having had seven different managers in the 1990s. Under its various managers, the fund has been bolder than its peers, focusing on lower-quality convertibles and riskier sectors, such as technology. Long-term returns are strong, but volatility is above average. This no-load fund carries one of the lowest expense ratios in the category.

Peter Di Teresa is a senior analyst with Morningstar. He can be reached at [email protected]