FXP Stock Trading

Some Reverse Engineering

Being an economist by training, I am inclined to seek frameworks to help me think about various issues. The context that I use here has its origin in observing the emerging markets trader Edward Cowen whom I worked with at Salomon Smith Barney in London in the late 1990s, as I mentioned in the Introduction. In working with him, I came to reference a model that I had previously resisted when I first came across it at the university—a model made famous by an analogy used by Milton Friedman, the Nobel-winning economist.

Friedman argued that the actions of a pool player may be largely predicted by a model constructed on the assumption that the player possesses high-level mathematical and other skills. These skills enable the player to compute quickly and accurately the angles and deflections that determine the best shot at any specific time. Friedman's approach was part of a more general, and highly controversial, debate about proper scientific method.3 There were strong proponents on both sides of the debate. Indeed, if one existed, the bumper sticker for this controversy would read something along the lines of "Whose side are you on: falsifiability or verifiability?"

Taking a cue from the work of Karl Popper, a highly regarded philosopher, Friedman argued that the best test of a model is to compare its predictions with actual outcomes. To this end, it was defensible to construct the model on the basis of simplifying assumptions (as in the case of the approach used for the pool player). While one may bicker with this approach—and I would still do so today—it is nevertheless useful for thinking about how to reverse engineer a type of behavior that leads to specific outcomes. And whether they know it or not, many of today's hedge funds that use "black box" models are effectively applying Friedman's approach.

Today, the question is whether we could reverse engineer the reaction of Edward Cowen, the Salomon trader, by assuming that, as a starting point, he was anchored by a systematic approach to analyzing market noise and reacting accordingly. This in no way denies the fact that, in reality, sharp "street smarts" drove Edward's ability to decipher signals within the market noise. Instead, it is an attempt to reverse engineer his behavior on the basis of an analytical model.

In doing so over the years, I have come to the conclusion that such an exercise would yield the following six key steps:

1. Identify the source of the noise that creates an unusual market dislocation.

2. Be disciplined in treating each episode of such noise as potentially containing important signals.

3. Assess the actual signal content through an evaluation based on the a priori modeling of the economic or market phenomenon.

4. Differentiate between factors that influence the destination and those that influence the journey when assessing the content.

5. After you have gone through this process, and not until then, you should actively pursue the views of the experts and the talking heads.

6. Be open to finding not only cyclical influences but also secular ones.