From the July 2006 issue of Entrepreneur

According to the mutual fund trackers at Morningstar, 900 to 1,200 funds go to the Great Beyond each year. They liquidate or merge themselves out of existence. This is what happens after the funds have to carry the increasingly heavy baggage created by their own longer-term performance. Sometimes it's simpler for investment companies to fold their cards and start a new fund from scratch.

The significance of all this is what's known as "survivorship bias." Obviously, the funds that die tend to have lousy track records. So when you measure the past performance of a group of funds against a benchmark (large-cap funds, say, against the S&P 500 Index), the funds look like they performed better than they actually did because the worst performers are stricken from the record. It's like saying the average NFL team had a 10-6 win/loss record last year, ignoring the teams with losing records.

How much does it matter? A study released this spring by Savant Capital Management found that Morningstar's published data inflated returns in 41 of its 42 categories by an average of 1.6 per-cent per year between 1995 and 2004. Morningstar didn't deny the problem--it said it will debut new databases next year that will take survivorship bias into account for category returns.

For the most part, individual investors don't have to lose sleep over this. Even if their aggregate numbers are overly rosy, services such as Morningstar and Lipper are helpful if you want to search for a single investment. A five-star fund is still a five-star fund, even if the category average isn't accurate.

But the survivorship bias numbers do give a nod to index funds in that eternal debate between passive and active investing. In his 1973 book, A Random Walk Down Wall Street, Princeton professor Burton Malkiel concluded that most investors have little or no hope of beating the market averages over time. It's cheaper--and easier--to buy a low-cost index fund with the bulk of your investing dollars. Consider that we've been looking at artificially juiced returns for actively managed fund categories all these years, and the argument in favor of indexing only gets stronger.

Scott Bernard Nelson is a newspaper editor and freelance writer in Portland, Oregon.