Let's Put an End to the Industry Standard 1, 3, 5, and 10 Year Comparisons


Around this time last year (i.e. shortly after 12/31/2013), we recall how five-year annualized returns of various mutual funds got a substantial boost not only from adding the good year of 2013 but from dropping the dismal year of 2008. The authors of this study titled “Past Performance May Be an Illusion: Performance, Flows, and Fees in Mutual Funds” introduce a new term to describe this phenomenon—the horizon effect. To illustrate just how large this effect can be, consider the S&P 500 Index which had 5-year annualized returns of 1.66% and 17.94% as of 12/31/2012 and 12/31/2013, respectively.

The authors of the study state that investors appear to be unable to differentiate between the new and stale information components of performance reported by mutual funds. As a result, stale performance chasing is amplified for funds which promote performance via advertising, and fund companies exploit this behavior by increasing their fees when they see increased in-flows from investors. The Economist recently published an excellent summary of the study titled “Mutton Dressed as Lamb” which is a reference to the advertisements of the mutual fund companies. As Americans, however, we can more readily identify with the idea of putting lipstick on a pig.

As one of the study’s authors (Raghavendra Rau) was quoted in the The Economist article, “The study paints a picture that’s not very pretty. The effect is not unlike a Botox operation gone terribly wrong.” In many articles such as this one, IFA has warned investors against the hazards of performance chasing, and most of the advertising for actively managed funds that we see emphasizes short-term returns.  It appears to us that investors who are influenced by short-term performance are afflicted with a form of the gambler’s fallacy. The 2002 Nobel Laureate in economics Daniel Kahneman and the late Amos Tversky concluded that these investors tend to believe in “the law of small numbers.” In other words, they tend to generalize from small amounts of data such as five or ten years of mutual fund returns. Kahneman and Tversky were alluding by way of contrast to the Law of Large Numbers which states that the average of a large number of trials approaches the true average as the number of trials increases.

IFA’s own study of the performance of actively managed funds (see here) indicates that even 20 years is inadequate to conclude the presence of true skill. With a typical standard deviation of alpha around 6%, a fund with a true alpha of 2% would require 36 years of returns before we could conclude that skill rather than luck is the explanatory factor. The t-stat calculator below shows how this works.

As much as we would like to get rid of the industry standard 1, 3, 5, and 10 year comparisons, we do recognize that they stem from the Securities Act of 1933 which was designed to protect investors in the wake of the crash of 1929 and subsequent downturn that wiped out so many of them. We at Index Fund Advisors will continue to do our part to educate investors about the importance of having very long-term data (50 years or more) when making investment decisions.