Skill vs Luck in Money Manager Choice. How Can You Decide?

Posted on April 6th, 2009 by admin in Quant

When you choose a new money manager or evaluate your current one, the most compelling, irresistible question is “How have you performed for me lately?” It’s too bad the answer tells you almost nothing. The point is to pick some one who will do well for you in future, and past results tell you very little. The research bears this out. You can, however, select well based on other criteria.

Even if a manager has a history of high ROI, his next year could be awful. These days, almost every manager has a rather terrible performance record for 2008. So there are probably lots of very highly skilled managers who performed horribly due to the massive economic collapse. Similarly, a poor history could precede a great year, e.g., because of learning. So it’s natural to want to pick a manager on the presumed likelihood of good future performance. But is ROI track record a useful predictor of skill or future performance? No.

Investment skill is a property of human mind, and thus cannot be specified or observed independently of its behavioral results, in this case, ROI. Thus, we must measure past investment performance and try to infer if some unobservable factor, called “skill” even exists. That still leaves open the question of how to detect skill in a particular manager.

One empirical approach to determine if skill exists takes a set of mutual funds with active managers (who rely on presumed skills) and compiles their performance over some time period. Through statistical  comparison with passively managed funds (non-skilled), researchers try to find explanations for the results that cannot be explained by “chance”, in which case the results are caused by an “X factor” loosely named “skill”.

Keep in mind that passive management requires the manager to track some reference benchmark.  Active management picks securities expected to have superior performance compared to the benchmark, but with no more risk.

The research results show that passive money management usually beats active, and that skill exists but is quite rare. (See the paper linked below, to review the details and subtleties of the research; emphasis added).

Luck_vs_skill_in_active_mutual_funds

The research still cannot predict who is the “most skillful” manager, and certainly cannot say which manager will have the best ROI next year.

An entirely new line of inquiry shows even persistent differences in ROI over time are not necessarily reliable indicators of differences in managers’ skills. Here’s the link (emphasis added):

Nickel-vs-black-swan-strategies

The differences may simply reflect differences in reputational concerns. Some managers may be so concerned to maintain their reputation they select an overly conservative strategy and perform worse than they otherwise could. Thus we now have two major factors, skill and reputational concern, that could account for a manager’s ROI results.

Obviously it’s hard to isolate the set of traits and reputational concerns sufficient to select a money manager. So, with all this difficulty isolating skilled managers, how can you choose?

First, ask yourself, and answer truthfully: Do I want to manage my own money? If yes, then do it. As your own manager, you will be picking securities and funds. You will necessarily decide for yourself  passive versus active.

If no, you must find another, even if you could do the job better yourself. This last point is really annoying, but it goes with the territory when you outsource. There are many things you pay others to do that you could do better yourself.

To select someone to manage your money, choose a manager who will wisely steward your money. Select based on your assessment that the manager will pay sufficient attention to your account and make your interests his highest concern. (The commitment to your interests is a fiduciary duty of Registered Investment Advisors, but not of stock brokers.)  He also should have the technical competence to handle various economic and market situations as they arise. In this context, “technical competence” is not the same as the delightful “investment skill” we all seek.

Technical competence can be observed, based on experience and education. Note than even passive management still requires portfolio adjustments from time to time, if for no other reason than to ensure meeting allocations or maintaining diversification. So the decision comes down to relationship building based on observable but subjective traits of commitment, competence and trust.



Leave a Reply

More News