An article in InvestmentNews discusses how “average investors” in some of the top ten large-cap mutual funds failed to beat the market over a recent five-year period due to market timing:

Over the five-year period ended Aug. 31, which includes the collapse of Lehman Brothers Holdings Inc. in 2008, the S&P 500’s 42% free fall to the bear market’s bottom and its subsequent 130% rally, five of the 10 biggest large-cap-stock funds posted better annualized returns than the benchmark.

These five funds are:

According to the article

The average investor return, which takes into account buying and selling behavior, for all but one of the funds was much lower because investors were busy selling, according to Morningstar Inc.
Only investors in the T. Rowe Price Growth Fund enjoyed the full market cycle’s outperformance. The average investor return over the past five years in the fund was 8.85%, beating the fund’s 8.63% return.

So, what’s wrong with the article’s thesis? First, a simplified proposition of looking at just the fund’s returns without taking the fund’s risk into account. Second, the use of a single performance benchmark (the S&P 500® index, a proxy for “the market”) that cannot fully adjust for that risk. Third, a notion that there exists an “average investor” whose investments into and withdrawals from the fund precisely mimicked the inflows and outflows generated by all of the fund’s investors in both time and relative scale (for one, there is no proof that these same investors who cashed out later reinvested into the fund). Fourth, a surprising indication that market timing may actually work in some cases (e.g. PRGFX). Fifth, looking at just one specific time period for a very small number of funds to derive the following speculation:

After all, five years from now, it may be funds such as the American Funds Growth Fund of America (AGTHX) or the American Funds Investment Co. of America Fund (AIVSX) that are sporting the best 10-year annualized returns, even though both have underperformed these past five years.

How would a truly risk-adjusted performance of the five funds look like over an extended period of time, which includes the interval used in the article? Here is one example, as a continuation of a previous Alpholio post on Fidelity® Contrafund®:

Cumulative RealAlpha™ for FCNTX

The chart clearly shows that on a cumulative RealAlpha™ basis, the fund started to underperform its reference portfolio in late 2007, i.e. well before the onset of the financial crisis. Therefore, the following five-year period was largely irrelevant to those investors who practiced “market timing” based on the above information.

Pin It