The authors examine the difference between mutual funds’ buy-and-hold, or time-weighted, returns and the average dollar-weighted returns, or IRRs, that are earned by end investors over the January 1991–June 2013 period. The question they seek to answer is why the value premium, shown to be persistent and large, remains an exploitable anomaly. Hsu, Myers, and Whitby find that value fund investors consistently underperform the funds in which they invest because of ill-timed buy and sell decisions. They assert that mutual fund investors are effectively financing the value premium and that their behavioral biases toward trend-chasing investment decisions are helping to ensure that it remains. Other reasons, in addition to performance chasing, may account for fund investors’ timing decisions, but the authors’ purpose is not to seek an exact explanation. Their interest lies in documenting the existence of the underperformance. Their findings of underperformance are consistent with results previously published by Jason Zweig (2002) and Russel Kinnel (2013).
Using the CRSP Survivorship-Bias-Free U.S. Mutual Fund Database as the source for monthly return and quarterly fund characteristic data, the authors create equity mutual fund portfolios weighted by total net assets. A fund is classified as being an equity fund if greater than 80% of its assets, on average, are allocated to equities. If a fund does not disclose its allocation to equities, it is classified based on its Wiesenberger, Strategic Insights, Lipper, or Morningstar classification. Funds with total net assets below $10 million are excluded from the data set. Size and style classifications are made using the fund’s prospectus-stated benchmark, obtained from Morningstar Direct and mapped to the CRSP data by fund CUSIP. The Active Share methodology of Cremers and Petajisto (2009) is used when a fund benchmark is not available.
Over the period analyzed, mutual fund value investors underperformed the funds they invested in by 131 basis points (bps) per year. The average value fund manager outperformed the buy-and-hold benchmark by 39 bps net of fees (unadjusted for style or risk), while value fund investors underperformed the benchmark by 92 bps.
But the value premium is not the only anomaly that is squandered by poor timing. Using the same methodology applied to the value funds, the authors find that investors in growth, large-cap, and small-cap mutual funds also underperform their funds’ time-weighted returns. Over the period analyzed, the average growth fund investor experienced the largest performance gap (316 bps a year) compared to the fund’s return. This is more than twice the gap associated with the average value fund investor. The gap for small-cap investors was 155 bps a year—the fund manager was found to outperform the S&P 500 Index by 81 bps, while the fund investor underperformed the index by 74 bps.
The authors also undertake a multivariate analysis of individual mutual funds to better assess the potential determinants of the return gap. To be included, a fund must have a minimum of 48 months of return data. Characteristics explored in the analysis include the log of the fund’s average total net asset ratio, the fund’s average expense ratio, and a dummy variable for a front-end load. Other dummy variables denote if the fund is growth, value, small cap, large cap, index, or institutional. Independent variables are the natural log of the total net assets of the fund, the fund’s expense ratio, and fund age. The results of the multivariate and univariate analyses are consistent. The growth and value indicators (the authors’ main interests) are significant with expected signs: the performance gap of value investors is smaller than that of growth investors.
Other multivariate findings include: 1) index fund investors have experienced a smaller gap on average than non-index fund investors, 2) institutional fund investors have realized a smaller gap than retail fund investors, and 3) investors in high expense ratio funds experience a much larger gap (401 bps a year) than investors in low expense ratio funds (134 bps a year). The authors conclude that the phenomenon of making poorly timed allocation decisions is observed disproportionately in less sophisticated investors, who are more likely to invest in funds with high expense ratios and who do not qualify for institutional share class funds.
The authors conclude by emphasizing that the return derived from a buy-and-hold implementation of an investment strategy—specifically, value and growth—is, more often than not, an inaccurate and overly optimistic estimate of value-add given the reality of fund investors’ suboptimal allocation timing decisions.
Summarized by Kay Jaitly, CFA.