Saturday, December 11, 2010

Unintended consequences

Performance attribution seeks to identify how a manager's decisions contribute to their returns; and while there is sometimes controversy about using at a model that analyzes things separate from these decisions, there can often be benefits as a result of seeing how unintended actions impacted the results.

No doubt Bernie Madoff never expected that his greed would result in the suicide of his older son, Mark, as reported today.Apparently the pressure of the investigations and at least one lawsuit was such that he felt there was no alternative. I have known individuals personally or through connections who took such action, and know that what they leave behind is usually a family that is devastated and stricken beyond imagination; no doubt this is happening with Mark's family.

We all have heard that his sons were apparently unaware of Bernie's crimes, and there is no way to know for sure, but I can see how this could have occurred. Regardless, I am sure that Bernie never expected that his actions would cause one of his sons to take such a drastic step. Bernie's actions have been tragic for so many, and here's just another sad example. My heart goes out to Mark's wife and children.


  1. David,
    Putting aside all the deep sadness of the Madoff affair, especially the suicides that have been associated with it, I want to reply to your opening paragraph and the technical point it addresses.
    I believe that there are two valid but emphatically distinct kinds of financial performance attribution whose conflation can only dangerously mislead and never be beneficial. One type, like market factor analysis, etc., explains how different market effects that are not controllable by the fund manager did or are projected to affect performance and/or risk. The other type, which I like to group under the term decision evaluation, explains how different investment decisions that are under the control of the fund manager did affect active properties regarding performance and/or risk. Of course, a market move can influence a fund manager’s investment decision and an investment decision can lead to unexpected consequences. However, the point of decision evaluation is to determine the total effect of a particular decision on a particular active performance property of a fund and that then will, by necessity, include the contributions of that decision’s ‘unintended actions’ on that property. But, since it intends to evaluate the effect of each decision, it does not include a separate evaluation of the effect of the unintended action on that property or on anything else. Also, the effect on aspects of the world other than the particular active performance property being addressed is not properly the purpose of any type of performance attribution. Furthermore, I believe that it is important to precisely define what one means by “how a manager’s decisions contribute to their returns.” For the evaluation of the effect of a market factor or a decision to be meaningful and useful, it must be constructed as a proper exact answer to an economically and precisely formulated question and not be approached informally or even as a purely formal exercise in decomposition.
    The reason that I believe it is important to be so careful about all this is that I believe that what is not correctly measured cannot be properly controlled or improved. And, as we found out with the Madoff affair, a lot of money, and even lives, can be at stake.

  2. Andre, thanks for both your insights and elaboration!


Note: Only a member of this blog may post a comment.