Finding Skilled Managers

There are a number of ways, including those proposed by Wermers and Surz, for identifying successful active managers who are pursing successful strategies. At AthenaInvest, we focus on the strategy being pursued by the fund and we have developed a portfolio building methodology we call Strategy Based InvestingSM. Using this methodology, it is possible to construct better active equity mutual fund portfolios.

My results reveal that it is easier to find skill among short history managers. Several respondents had trouble with this counterintuitive result, so let me explain why I believe this is the case. Performance declines over time because the manager, once successful, comes face to face with the performance killing forces of the industry. These include growth in AUM, marketing pressures such as having to stay in a box (e.g. small-cap value), having to meet short term budget numbers, reporting to an investment committee, responding to requests for lower volatility, good upside capture, and low downside capture, and so forth. The latter is the unreasonable expectation among advisors and investors alike that somehow a manager can generate superior performance while also producing consistent short term returns. The evidence, both at the market and individual security level, shows that long term superior performance is accompanied by highly unpredictable short term returns and thus eliminating short term volatility has the undesirable consequence of hurting long term performance.

This is not to say skill cannot be found among long history managers. On the contrary, one would expect skill to increase with experience, but the performance killing forces of the industry grow ever stronger the longer the fund is around. It takes a tremendous force of manager will to push back and stay focused on strategy. This is particularly the case as fund organizations increase in size with the attendant growth in staff and the pressure to meet short term budget numbers. And unfortunately, advisors contribute to this pressure. Often I hear advisors say they are more interested in short term consistency than in long term performance. How sad for their clients.

Manager search methodologies, such as those offered by Surz, can help get beyond the standard performance numbers. Many advisors have their own method for identifying skilled managers that can help in building better portfolios. Often these correctly focus on the strategy being pursued by the manager and how successful the manager has been in executing the strategy.  Excluding managers with shorter than a five year performance record is a mistake and focusing solely on a long term track record is fraught with problems, as the empirical relationship of declining returns with longer history reveals. The search process is more complicated than simply looking for a long term track record.

Methodology Questions

There were several comments regarding the use of correct benchmarks. If one is talking about individual fund performance, then the S&P500 is not the appropriate benchmark and a carefully selected benchmark is needed. But since I present aggregate data over thousands of US equity active mutual funds, hundreds of thousands of fund/month observations, and a nearly 30 year time period, using the S&P500 benchmark is appropriate. There is no reason to believe that over time mutual funds as a whole change, relative to the market, along such dimensions as sectors, market cap and P/E, and so it is unnecessary to make additional adjustments.

Risk models, such as the Carhart four factor model used by BSW, adjust for relative risk among funds but not aggregate risk. This means they adjust for risk differences from one fund to another, but do not adjust for increases or decreases in collective fund risk relative to the market. Since I report aggregate fund results, such risk adjustment models provide no insight into the overall risk being taken on by funds that might help explain the observed increase in average alpha.  Add to this the large data cost (consuming as much as 50% of the sample) of such models, and I conclude the costs outweigh the benefits. In this sense, the average standard deviation I report is an adequate measure of changes in aggregate fund risk. And since the average fund standard deviation declines over this nearly 30 year time period, changes in aggregate risk cannot explain rising alphas during the 1980-2008 time period.

Building Better Client Portfolios

The current market environment favors active management over indexing and there are tools available to help find skilled managers. My hope is that advisors take advantage of the current opportunity to build better client portfolios.

Display article as PDF for printing.

Would you like to send this article to a friend?