(MONEY Magazine) – When you peruse the mutual fund rankings that begin on page 140, please first take a moment to enjoy the view. During 1997's first six months, the markets once again treated stock and bond fund investors generously. For that period, the hottest stock fund, wildly volatile American Heritage (page 140), was up a delicious 78.6%. The top bond fund, Phoenix Emerging Markets Bond A (page 142), gained a gratifying 16.1%. Even the average diversified U.S. stock fund returned an impressive-sounding 13.2%.

But before you send roses to your fund manager, consider this: The benchmark Standard & Poor's 500-stock index was up even more--20.6%. Only 5% of all U.S. stock funds beat the S&P; just 21% managed to whip the broader Wilshire 5000 index. For a longer-term perspective, look at the table on page 139 and you'll see that the average stock fund has lagged the S&P for the past one, three and 10 years as well. Worse, those averages overstate the annualized performance of funds by 0.14 to 0.68 percentage points, for reasons I'll explain in a moment.

The central lessons of these performance comparisons are abundantly clear:

--First, long-term investors should make low-cost, market-matching index funds the core of their stock and bond fund portfolios. Wish you had beaten the vast majority of mutual fund owners (95% so far this year and 77% over the past 34 years)? Then buy index funds, because their superior performance, in relation to other funds, looks likely to persist for the long term. We've made the case before, and while index funds won't protect you from losses in the kind of temporary market slide that Money's Michael Sivy predicts (on page 68), we're here to issue that same long-term advice, supported by new, ground-breaking research.

--Second, in today's number-happy mutual fund marketplace, you need to interpret fund performance figures--especially those that show a fund category's long-term average gains--with a new sophistication. You must think about those three-year-or-longer aggregates, including the ones you see in Money, the way the pros do: skeptically. That's because the numbers can mislead. Or, if you employ them in the ways that I'll explain, they can make you a better-informed and more successful investor.

So how exactly do fund category averages lie? I doubt that anyone has ever told you this before, but the "average" return for a category of funds--whether a large subset like diversified stock funds or a narrower one like small-company growth--tracks only the performance of the portfolios that survived all the way from the beginning of the measurement period to the end. But in a typical year, about one fund in 28 kicks the bucket. Each time that happens, the rating services like Lipper Analytical and Morningstar expunge the dead fund from their historical averages, wiping away that fund's returns for all previous years.

Guess which funds tend to go kerflooey? The ones with poor performance records. And though the rating services compute their category averages as if these funds never existed, they of course lived real lives and collected money from real shareholders who earned real returns (or, should I say, real bad returns). I refer to these as zombie funds: Though they are deceased, their often horrifying performances continue to cast a shadow over the fund tables.

To understand the distortion this causes, imagine that you did a survey of retired race-car drivers and found that their average age was 86. Would you conclude that screeching around racetracks at 180 mph is a good way for people to prolong their lives? Of course not. No one would ignore the drivers who died young in a bloody blaze of fire. The average age of the survivors might be high, but the average life span of all drivers from that era, living and dead, would be a lot shorter. This statistical trap is called survivor bias, and the term is as fitting for funds as for auto racers. The long-term "average" returns of fund categories, as now computed, ignore all the real-world investors who buy funds that crash and burn.

Morningstar, which provided the data for MONEY's fund tables, "is committed to purging survivor bias" from its databases by the end of this year, says Don Phillips, the company's president. But there's no need to wait. I'm about to give you the tools to compare accurately the returns of different types of indexed and managed funds.

In a paper so fresh it hasn't even been published, Mark Carhart, an assistant professor of finance at the University of Southern California, precisely documents the havoc that zombie funds wreak on "average" category returns. Other researchers had already shown that survivor bias can skew fund averages, but Carhart has done them one better. He has painstakingly constructed a database of all diversified U.S. equity funds (at least those for which any records could be found) that existed at any time between January 1962 and December 1995. And he has demonstrated how investors suffer earnings shortfalls in funds that decline and die. Here's what Carhart found:

--Over the 34-year period he examined, out of a grand total of 2,071 stock funds that were ever open for business, 725 disappeared. Put another way: 35% of all stock funds went kaput, for an average annual mortality rate of 3.6%.

--Counting only the survivors, as most performance tables do, the "average" fund returned 10.7% compounded annually from 1962 through 1995. That suggests the typical fund manager was at least earning his keep, since the S&P 500 index returned "only" 10.6% compounded annually over the same period. But when Carhart included zombie funds, the average compound return for stock funds dropped more than a percentage point, to 9.5%. That dramatically transforms the typical fund from a market beater to a market laggard. (The S&P index, by the way, suffers almost no survivor bias, mainly because very few stocks drop out of the index for reasons of poor performance and because its average annual returns are not restated every time the membership of the 500 changes.)

--The extra return promised by aggressive growth funds is much smaller--and far riskier to chase--than you've been told. The raw "averages" say that from 1962 through 1995, the typical growth and income fund returned 10.8% compounded annually; the typical growth fund, 11.5%; and the typical aggressive growth fund, a whopping 13.5%. But after Carhart adjusted the averages by including the zombies, he found that growth and income funds gained 10.4% compounded annually; growth funds, 10.5%; and aggressive growth, 11.6%. (Over the same period, remember, the S&P 500 returned 10.6%.)

Now, you might conclude from this that your best hope of beating the market is still through aggressive growth funds. And you might be right--if you bought a dozen or more of the things and held them for at least 10 years. Trouble is, most of the investors I know who go for aggressive growth funds buy one or two of them. And that's dangerous, because the flameout rate among funds in this category was the highest of the major groups Carhart studied. Fully 39.4% of 614 aggressive growth funds, or 4.5% a year on average, crashed between 1962 and 1995. By contrast, just 33.5% of 827 growth funds and 32.7% of 630 growth and income funds hit the wall.

Those 725 zombie funds that Carhart identified were scary in more ways than one. Over the final five years of their existence, they underperformed the average fund by a cumulative total of more than 20 percentage points.

I think Carhart's work offers five valuable, practical lessons for investors:

--First--and again--buy index funds for the long term. If we consider only the survivors, 29% of funds beat the S&P 500 over the 34-year period Carhart studied. But among all funds, living and dead, only 23% outperformed the index; 77% did not. The message: Buy an index fund and, history suggests, you've got a better than 3-to-1 chance of outperforming actively managed funds. Because managed funds charge higher expenses than passive, computer-guided index funds (and trade their stocks more rapidly, churning up a bigger tax bill), it's tough for the active managers to beat those boring-but-beautiful indexers.

Now, to be fair, as Michael Sivy points out, it is precisely the large, blue-chip stocks in the S&P 500 that have gained the most in the bull market's latest surge, which makes them vulnerable now. Meanwhile, the smaller stocks favored by the stewards of actively managed funds have been underperforming. The stock market does move in cycles, and smaller stocks will bounce back. Then active managers will shine, at least for a time, as they did in 1991, when 53% of actively run U.S. stock funds beat the S&P 500.

Still, my long-haul choice is the indexers. If you want to gamble that big stocks will continue their hot streak well into the future, you can put the bulk of your money into an index fund that takes the beefy S&P 500 as its benchmark. Two superb choices in this category are Vanguard Index 500 (up 20.5% in 1997's first six months) and Fidelity Spartan Market Index (up 20.4%). But if you are concerned that big stocks are overvalued, consider a broader portfolio such as Vanguard Index Total Stock Market (17.6%), a fund keyed to the Wilshire 5000 index of virtually all publicly traded stocks in the U.S. With a total-market index fund, you'll benefit proportionally when small stocks bounce back, since they're automatically included in the portfolio package.

What about index funds in other markets? Active managers of international stock funds are, in my opinion, a little more likely to beat their benchmarks than U.S. fund managers are. Over the past 10 years, for instance, 52% of all diversified international stock funds have beaten Morgan Stanley Capital International's Europe, Australasia, Far East (EAFE) index of foreign stocks. But the zombie fund problem means that we have lost track of the international fund managers who have failed. And thus the case for international indexing too is better than you might think. Here, a good umbrella approach is Vanguard STAR Total International (up 11.7% for 1997 to July 1), which seeks to follow the performance of the EAFE index and its Select Emerging Markets Free index.

--Second, don't use recent fund performance as a guide to picking long-term winners. In my opinion, Carhart's research counsels skepticism toward those studies that claim you could have beaten the market over time by investing in the hottest funds of the previous year. This so-called hot-hands theory relies on the high returns of funds that were hot for awhile and survived for at least a few more years. But it ignores the returns of funds that quickly went from sizzle to fizzle. In the real world, if you had chosen your funds from among each year's stars, you would have bought lots of funds that expired.

To see what I mean, consider this: If you had invested in the hottest fund of 1987, it would have been Oppenheimer 90-10, which gained 93.5%. But that fund lost 4.9% in 1988 (while the S&P 500 rose 16.6%) and disappeared with its merger into Oppenheimer Asset Allocation in 1991. Similarly, the top fund of 1991 was Oppenheimer Global Bio-Tech, up 121%. It lost 23% in 1992 (the S&P was up 7.6%) and was absorbed into Oppenheimer Global Emerging Growth in 1994. As you can see, chasing recent hot performance is a good way to get burned.

--Third, approach aggressive growth funds as you would any dangerous, volatile creature. When you count the dead ones as well as the live ones, the aggressive funds barely outperform stodgier types of funds--but incur much more risk. Instead, you should keep the bulk of your stock assets--at least 75%--in broad-based index funds or large-stock funds, where, over the long run, you'll get similar returns at lower risk.

--Fourth, never decide which kind of fund to buy based on the long-term "average" returns of various categories. These peer-group averages are almost always inflated by the zombie fund factor--and the longer the time period, the more the averages are likely to be overstated. Average returns for fund categories, like those in the table at right, are useful in showing how your individual fund stacks up relative to its category peers. But such figures are not a reliable yardstick for measuring how much of your total portfolio you should devote to a particular category of funds. Instead of picking a fund just because it's in a category with high average returns, you should choose either an index fund, or a managed fund that has a clear and simple strategy, consistently good (but not spectacular) long-term returns, low portfolio turnover, high tax efficiency and economical annual expenses. Some excellent choices that meet these criteria include Fidelity Growth & Income (three-year return 26.1%), Scudder Growth & Income (25.1%) and T. Rowe Price Equity-Income (24.8%).

--Finally, whenever you see high reported "average" returns, you should conclude that the risk of fund death or disability is also higher than average. Don't let yourself be lured into chasing market-beating returns without first considering whether you can stomach sharp swings in value and whether you have diversified your holdings sufficiently to offset the risk that some of your go-go funds will end up goners.

As you read the tables of fund performance on the following pages, bear in mind that they are a terrific way to update yourself on how selected funds have performed so far in 1997 (page 140) and over the past one, three and 10 years. Here you can see how your funds fared or how they rank against the competition, just as any sports fan does when looking up the league standings in the sports pages.

But don't look only at raw performance. Instead, pay particular attention to two columns of data: annual charges and tax efficiency. Performance comes and goes, but the expenses a fund charges you to run your money are pretty much constant--and are one variable you can control up front, through smart shopping. The more the fund manager charges, the less you get to keep. In general, I think it makes sense to limit yourself to U.S. stock funds with expenses under 1% a year (1.25% if they buy small stocks), foreign stock funds that charge less than 1.5% and bond funds that cost 0.75% or less.

Likewise, a fund that frequently buys and sells stocks is more likely to produce taxable short-term capital gains and give more of your earnings to Uncle Sam. Instead, look for funds with high "tax efficiency," clearly identified in the tables that follow. (For funds that you own in a tax-advantaged plan like an IRA, Keogh or 401(k), tax efficiency is not an issue, since all of your fund returns can compound tax deferred.)

As any baseball fan knows, yesterday's box scores and today's league standings don't tell you everything you need to know to predict which team will ultimately win the World Series. It's certainly exciting if one of your funds is in our top 100 table for 1997 performance. But the more you focus on short-term results, the faster and further the long term will run away from you. If you're shopping for a new fund, you should choose from among the long-term winners covered in the tables that begin on the next page.