Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
Support MFO
Donate through PayPal
The Harm In Selecting Funds That Have Recently Outperformed: Research Paper
FYI: We empirically investigate the investment results of commonly used fund selection strategies that involve redeploying assets from underperforming to outperforming funds. Based on portfolios constructed using U.S. mutual fund data over typical three-year evaluation periods, we find that investors who chose funds with poor recent performance earned higher excess returns than those who chose funds with superior recent performance. Our findings pose a challenge for asset owners: If past performance is used at all in selecting funds, it is the best-performing funds that should be replaced. Realistically, however, a policy of replacing successful funds with poor performers is unlikely to gain widespread acceptance. Instead, the practical implication of our paper is that asset owners should focus on factors other than past performance. We offer alternate criteria for selecting funds. Regards, Ted http://poseidon01.ssrn.com/delivery.php?ID=438094071086119066002010083088101123024020030032038022077085101018094027090108104009120036123104050034053099090030004100110109109040002033054075065068084123004103058015033101012027103091100086127029094104004097087071074014106113086015101099123013103&EXT=pdf
I went right to the results of this paper. Everything else puts me to sleep.
Results from the this investigative paper:
…a strategy of hiring managers with mediocre track records outperforms one of hiring past winners, and a strategy of hiring past losers turns out to be the best of all.
…the practical implication of our paper is that asset owners should focus on factors other than past performance when selecting managers.
…the “investment thesis” that drives a fund’s portfolio management strategy should be a key criterion for consideration.
…a variety of objective characteristics that predict future performance… :the presence of performance-linked bonuses in fund manager compensation packages (Ma, Tang, and Gómez (2015)), a high level of fund manager ownership (Khorana, Servaes, and Wedge (2007)), board of director ownership (Cremers, Driessen, Maenhout, and Weinbaum (2009)), a high active share (Cremers and Petajisto (2009), Amihud and Goyenko (2013)), lack of affiliation with an investment bank (Hao and Yan (2012)), outsourced execution of shareholder services (Sorhage (2015)), the presence of a short-term redemption fee (Finke, Nanigian, and Waller (2015)), having PhDs in key portfolio roles (Chaudhuri, Ivkovich, Pollet, and Trzcinka (2013)) and having strong positive firm culture (Heisinger, Hsu, and Ware (2015)).
In conclusion: Evaluating a manager’s strategy ex-ante and taking account of fund characteristics may be more difficult than making decisions based on historical performance. Nonetheless, our research suggests that it is a better approach to delegated portfolio management.
Much if not all of these results have been stated by different people here at MFO, which makes this site so enlightening. It strengthens my resolve that there are only a few places where you might hire a fund manager. For me, International, EM's maybe small caps need active fund managers. Balanced funds for sure if you want to leave it up to a professional to adjust investment weightings (which I do).
I looked at this study but could not figure out whether the authors' data suffer from survivorship bias. If the dead funds are not included in the database and analysis, the findings of this paper could mean nothing. Suppose that a fund disappeared due to poor performance in year t. If this fund was kept in the database, a strategy of including it in the loser portfolio at t-1 or t-2 would generate poor results. From a strategy implementation point of view, this makes sense because at t-1 or t-2 an investor would not know that this fund will be liquidated at t. And given that there is a large number of dead funds, the strategy of picking loser funds would end up picking funds that fail.
What makes a good fund manager? One who consistently beats his index. Very hard to do, of course, but once done then quite clearly such a manager will have performed better than simply rotating to the worst mangers serially. The study did nothing but show expected reversion to the mean, which I think is the most powerful factor in investing next to compound interest.
I looked at this study but could not figure out whether the authors' data suffer from survivorship bias. If the dead funds are not included in the database and analysis, the findings of this paper could mean nothing. Suppose that a fund disappeared due to poor performance in year t. If this fund was kept in the database, a strategy of including it in the loser portfolio at t-1 or t-2 would generate poor results.
"If a fund [from t-1 or t-2] disappears from our dataset [in year t], then the capital that was invested in it is equally allocated among the remaining funds" in its cohort, i.e. top decile, middle decile, bottom decile.
Yes, dead funds are included. Note that they also excluded the 10% most expensive funds, since high costs drag performance down. Why 10% you ask? So do I.
Thanks, Msf! Not the smoothest read and I was simply looking at the data section. Also, I agree that the 10% filter does not make any sense and appears arbitrary. I wonder whether these results still hold without this filter.
Comments
Results from the this investigative paper: Much if not all of these results have been stated by different people here at MFO, which makes this site so enlightening. It strengthens my resolve that there are only a few places where you might hire a fund manager. For me, International, EM's maybe small caps need active fund managers. Balanced funds for sure if you want to leave it up to a professional to adjust investment weightings (which I do).
@MikeM- Yes! Yes! Hire me! Hire me!!
Yes, dead funds are included. Note that they also excluded the 10% most expensive funds, since high costs drag performance down. Why 10% you ask? So do I.
Also, I agree that the 10% filter does not make any sense and appears arbitrary. I wonder whether these results still hold without this filter.