Mutual Fund Returns: Not Always What They Appear

You might not have noticed it, but you've probably never seen an advertisement for a mutual fund that reports a failing return. It's an impossibility that every fund could perform so well all the time. But what happens to those inevitable lemons if the mutual fund industry denies any are falling? This is where survivorship bias steps in.

A mutual fund company puts survivorship bias into action when distorting the true performance of its mutual funds, making the funds look more attractive than they really are. This bias is created when poor-performing funds are liquidated or merged into better-performing funds. As a result, these substandard performers, and their corresponding substandard metrics, simply disappear.

When these "losers" are purged from their respective categories, their statistical records are no longer included in the category performance data. This makes the category averages creep higher than they would have if the losers were still in the mix.

Example: The Survivorship Effect Let's say that there are three funds (A, B and C) in a given category. Fund A has a five-year annualized total return of 12%; Funds B and C have five-year annualized total returns of 8% and 4%, respectively. The average annual total return for the fund category would be 8%. But, if the loser, Fund C, were to be liquidated or merged into either A or B, it would disappear from the data radar screen. The five-year average annual total return for the fund category would become a more impressive 10%.

Hedge funds can also fall into perils of survivorship bias. Many research and database firms, however, didn't start collecting data on retired hedge funds until 1994, so research on this area has yet to come to a definite conclusion. Just be careful when looking at any hedge fund returns reported before 1994 because there is a good chance survivorship bias skews the numbers significantly.

Fund companies argue they shouldnt have to include dead funds in return calculations because the funds are transferred to different managers. But think of it this way: When a person buys a new car he or she doesn't get to erase all past accidents and speeding tickets, so why is it that mutual fund companies virtually get to erase their past mistakes?

The CFA Institute (formerly the Association for Investment Management and Research) has attempted to place restrictions on how past performance is reported. However, this disclosure isn't a requirement for mutual fund companies. They follow the restrictions only if they choose.

Even companies that do usually comply only need to publish their true performance in the fine print on the prospectus or other promotional material that most investors don't read. The Financial Industry Regulatory Authority and Securities and Exchange Commission have also made decisions on how funds report their returns, but there is still a gray area that can be (and often is) exploited by many companies. (See also: Footnotes: Early Warning Signs for Investors, An Investor's Checklist to Financial Footnotes and Footnotes: Start Reading the Fine Print.)

Creation bias is another form of survivorship bias. Creation bias works by giving a handful of investment managers a small amount of money to incubate their own funds. After a couple of years, the fund company chooses the manager who has performed the best. The successful funds are then made available to the public and marketed aggressively, while the losing funds get silently discontinued. Many investment professionals believe that creation bias is becoming a bigger problem than survivorship bias particularly because it is much more difficult to detect.

John Bogle, founder and former chairman of the Vanguard Group, often cited the survivor bias phenomenon as one of the reasons for favoring index funds, which don't play the survivorship game. Bogle is quoted as saying that "what we are really looking at here are 'juiced' managed fund performance numbers, which create a misleading picture that actively managed funds are competitive with indexing."

Survivorship bias is a kind of grade inflation for mutual funds that occurs when the funds with the worst performance are made to disappear from the database while strong performers survive another day. The result of this mutual fund Darwinism results in skewed performance numbers that make the remaining active managers look better as performers vanish before they can drag down the overall performance numbers of the category index.

The issues of survivorship and creation bias demonstrate the importance of being skeptical of mutual fund performance claims, particularly when the claims are coming from the company itself. The key, as in so many cases in investing, is to do your research.