Dead Man's Curve

Tavakoli Structured Finance, Inc.
By Janet Tavakoli - September 21, 2006 [HedgeWorld Sept 22, 2006 with permission of TSF]

Let's hear it for brain damage. The Journal of Economic Literature reported that with regard to investments "frontal damage can result in superior decisions," in the article "Neuroeconomics: How Neuroscience Can Inform Economics" in its March 2005 issue.

At least that is what the authors—Colin Camerer of the California Institute of Technology's Division of the Humanities and Social Sciences, George Loewenstein of Carnegie Mellon's Department of Social and Decision Sciences, and Drazen Prelec of the Massachusetts Institute of Technology's Sloane School of Management—believe, even though such damage results in poorer overall decision-making ability.

The idea is that frontal damage makes one more willing to accept reasonable risk than those of us without brain damage. For example, those with frontal damage will accept a 50:50 bet in which they will either win $300 or lose $200; whereas most people of sound mind will not accept the bet until they have the possibility of winning twice as much as they may lose: the 50:50 chance of winning $400 versus losing $200.

The authors use this example of aversion to loss to suggest that those with brain damage might be better investors. I am not making this up. To try to infer these people will make better investors is simply a false analogy, because the market does not present us with certain [known probabilities with fixed] outcomes. The reality is that this isn't a "superior decision," it is simply a greater willingness to accept a lower margin of safety. In fact, this behavior is labeled an "inferior decision" when the market tanks and one hasn't managed the risk.

What the authors failed to take into account is that taking more risk when upside is skewed (for instance, by buying on margin in a bull market) will usually produce better results. The true test of a good investor is when there is the possibility of unacceptable loss or when one is dealing with a market in which outcomes are uncertain [unknown probabilities with variable outcomes]. Brain damage is most unhelpful in these circumstances, since what seems like a certain upside is merely an illusion.

Amaranth Advisors provides an object lesson. Just this August, Nick Maounis, Amaranth Advisors' founder and chief executive, claimed that Brian Hunter, his star natural gas trader, was very good at taking controlled risks. The reality was very different, even though the Amaranth trades seemed logical on the surface. Treasury traders often go long the current long bond (the 30-year, now the 29.5 year) and short the penultimate, off-the-run long bond. Among other trades, Amaranth was short fall natural gas futures contracts and long winter natural gas futures contracts in sequential years from 2006 through 2009 (according to CNBC).

But just as too much money flowing into these trades can collapse spreads in the treasury market, too much leveraged money flowing into the much thinner commodities market undid Amaranth's trades. [If we use the model for a normal (Gaussian) distribution, a five-standard deviation credit event should only happen once in every 7,000 years. Spreads tightened by five to ten standard deviations in the September/December natural gas spreads depending on which time period you use for your data.] In September, Amaranth had lost half of its value, skidding from $9 billion to only $3 billion in assets (according to the Wall Street Journal), having put on the classic "Dead Man's Curve" trade ("The last thing I remember, Doc, the market started to swerve…."). These trades stubbornly refuse to follow history's pattern.

Nick Maounis is an alumnus of Paloma Partners. Another Paloma alumnus, Nassim Taleb, wrote a book called Fooled by Randomness, with a new approach to the well worn territory that human beings are not good at assessing probabilities without formal training. We have a tendency to explain random events as if we had foreknowledge of the outcomes. Many hedge funds' success (or failure) in the market is the product of lucky (or unlucky) bets. If the bets randomly pay off, the lucky fund manager incorrectly believes he or she is a genius. This may explain Mr. Maounis' willingness to give his star trader so much rope.

Ironically, both Mr. Taleb and Mr. Maounis misapply probability theory. They both ignore conditional probabilities. Mr. Taleb suggests we give undeserved accolades to Warren Buffett and damns Mr. Buffett with faint praise: "I am not saying that Warren Buffett is not skilled; only that a large population of random investors will almost necessarily produce someone with his track records just by luck." [Emphasis in original].

But if Mr. Taleb wanted to use an example of success due to random luck, he should have looked to Amaranth's ephemeral success [Note: The second edition of Taleb's book was published well after Amaranth was founded; as I and others (Christopher Fawcett, head of Fauchier, specifically warned in advance that Amaranth's collapse was foreseeable) had noted prior to its implosion, its business model was indicative of transient success due to random luck]. If not Amaranth, any number of examples would have been more apt than Warren Buffett. If one does use Warren Buffett as an example, however, then it is remiss not to mention conditional probabilities.The reality is Mr. Buffett puts in a lot of hard work to uncover a margin of safety whenever possible. What is the probability of a successful investment, given that one has a sound method for analyzing a business? It is much better than the probability of success without the sound methodology, and the probability of disaster is very low. In contrast, a one-sided hedge fund bet presents an altogether different conditional probability. What is the probability of a disaster, given that one has merely leveraged a market bet based on historical relationships? If one is lucky one will do well, but if one is unlucky, the probability of disaster is about 100%.

This article was published in Lipper HedgeWorld on Sept. 22, 2006 with the permission of Tavakoli Structured Finance, Inc.