(MONEY Magazine) – WHEN YOU'VE GOT IT, FLAUNT IT. AND MUTUAL FUNDS HAVE got it--flashy returns following a year when stocks shot up 37.5% and bonds 18.5%. Now comes the flaunt. Thanks to those numbers, many publications, including Money, are bursting with mutual fund ads that brag about top returns.

One day this January, three funds ran big advertisements in the Wall Street Journal declaring themselves "No. 1" by some measure. The February 1996 issue of Money had 48 mutual fund ads, 33 of which (or 69%) boasted about high returns; by contrast, our February 1995 issue carried 37 fund ads, of which only 17 (or 46%) bragged about returns.

Let me be clear: There's nothing improper about these ads. They all conform with the National Association of Securities Dealers' disclosure rules. But let me be equally clear on this: You should take them with a shovel of salt.

Why? For one thing, Dec. 31, 1995 was a big day statistically: 1990 dropped out of funds' five-year track records, and 1995 took its place. A bad year (stocks lost 3% in 1990) was erased by a great year (1995 was the best year for the S&P 500 since 1958).

Abracadabra! Suddenly most funds' five-year records look a lot rosier--not because the funds are any better, but because they get to dump a downer year and add an upper. Over the five years from 1990 to 1994, the stock market returned an annualized average of 8.7%; from 1991 through 1995, however, stocks returned 16.6% annually, nearly twice as much.

Moreover, even tiny shifts in a measurement period can make a big difference in returns. One example: For the five years that ended Sept. 30, 1992, Keystone America Hartwell Emerging Growth had a 10.7% average annualized return. Just one month later, the fund's five-year return more than doubled, to 21.8%. Had the portfolio manager become a genius in one month? Of course not; the passage of that month erased the crash of '87 from Keystone's five-year record.

Note too that when the markets shoot up as high as they did in 1995, nearly every fund looks great. Among the 1995 returns that funds have been showing off in recent ads: 30.9%, 35.7%, 35.9%. Next to the stock market's 10.5% annualized average return since 1926, those numbers sound spectacular. But they're actually below average, since stocks rose so much last year. Even a bad jockey can look good on a fast horse.

One other point to ponder: Chances are, not many investors owned the funds for the full term and therefore earned the returns that the ads are trumpeting. Performance numbers are a sampling, not a full measurement, of returns. They do not represent the experience of most buyers, sellers or owners.

Why not? The one-year returns now being advertised assume that you made a single investment on Dec. 31, 1994, reinvested all your dividends and capital gains, then sold your entire stake on Dec. 31, 1995. Likewise, the five- and 10-year advertised returns assume you bought in one lump sum on Dec. 31, 1990 and Dec. 31, 1985 respectively--and closed your account at the end of 1995 (paying no taxes on any gains).

Of course, real life is nothing like that. Nobody owns funds for exactly one, three, five or 10 years; instead of doing every dollar of their fund business on New Year's Eve, investors buy (or sell) on St. Patrick's Day, April Fool's Day, Elvis' birthday or any other day when the fancy strikes them. What's more, many people invest over time instead of in one lump sum. Lots of folks take their dividends in cash. And nearly everyone has to pay taxes.

Let's take a real-world look at $14.6 billion Twentieth Century Ultra, the Kansas City-based fund that traffics in fast-growing mid-size companies. Roughly half of Ultra's 840,000 account holders first bought the fund in 1991 or 1992 (lured, most likely, by Ultra's 86% gain in 1991). From January 1992 through last October, Ultra averaged a 14.3% return, while the market returned 12.1% annually. Yet according to Twentieth Century, many of Ultra's 1992 buyers sold before the end of the period. Thus these itchy-fingered folks earned just 10% annually. Although they bought one of the hottest funds around, these people were just passing through, so they shortchanged themselves.

What should you do about all this? Advertisements can certainly point you toward top funds, but you've got to look behind the numbers. Remember that a year like 1995 can make even bad funds look good. Make sure you note whether the advertised time periods include unusually good or bad years in the market--or for the sector in which the fund specializes. In the prospectus and annual report, look at the fund's year-by-year returns as well as its long-term average returns; averages can paper over some ugly holes in the past. And focus on funds with solid records over long periods, not just last year's white-hot performers. There are a lot more lucky sprinters than lucky marathoners.