Trusting reason

In the last post I talked about the Prisoner's Dilemma "game" and the question of trust. This morning I read an article in The New Yorker that looks at the financial meltdown as "rational irrationality." 


One of the article's points is that almost all the players in the debacle of sub-prime mortgages and CDOs knew that it was unsustainable, but the way the game was set up they had to opt for short-term gain and rely on their knowledge to get themselves a chair when the music stopped. Well, we know how that worked.


The article also discussed the issues involved in the prisoner's dilemma, specifically that a key factor for each prisoner is trying to guess how the other prisoner is going to act (especially in the single-iteration version, where there is the possibility of a high penalty for dual defection). The role of the investor, like the prisoner, is not to see the world as it really is, but as he thinks others see it. What's the harm in that?


It made me think back to Long-Term Capital Management, a hedge fund and investment company founded in the early 90's by John Meriweather. LTCM used not only seasoned traders, but some extremely high-powered mathematicians, Myron Scholes and Robert C. Merton, who in fact went on to share a Nobel Prize. They were "quants" who had developed equations that helped LTCM discover and take advantage of certain discrepancies in the arcane realms of fixed income arbitrage and foreign government bonds. The rate of return on each trade was tiny, but leverage allowed them to turn a tidy profit. For a few years their annual return was in the area of 40%.


The point of this is that LCTM knew that their advantage depended on using an approach that only they used. They had discovered an opportunity where the information (equation) available to them was asymmetrical to anyone else's knowledge. And they knew that when enough other people figured out what they were doing, the advantage, and the profits, for LCTM would disappear.


So, of course, as the market caught up to them, they kept on doing exactly what they had done. Of course, there were other factors, such as the Asian collapse and Russian crisis, but then, their trades were affecting those very markets. In September of 1997 their value fell from $2.3 billion to $600 million (for the same portfolio).


This led to a bailout of close to $4 billion organized by the Fed. (Can anybody say the word "counterparties?") Eventually, all the banks that participated by purchasing pieces of the portfolio realized at least small profits. LCTM eventually went out of business but Meriweather started another hedge fund, JWM Partners, supposedly having learned his lesson about leverage. That knowledge was not enough for the recent crisis, and JWM went bust in July of 2009. Some lesson.


My point is that, on the one hand, we trust what we know. The killer equation is always going to be the killer equation. And, on the other, we may think we can gauge what is in the other prisoner's mind and change our thinking accordingly based on our hard-earned experience. Unless our own perceptions (and killer equations) get in the way, because that's what we've learned to trust.


I suppose we are left with the moral that "rational irrationality" is more a tautology than an oxymoron. Many of us may try to be rational in the way we deal with the world, but but we just don't always know enough about how the world is really working, so we keep doing what worked in the past. 


The only rationality is to accept the irrationality -- the asymmetry of information -- as a fundamental fact of existence, closely tied to constant change. Even the smartest quants need to be reminded that the world is more complex, and more changeable,  than any equation. And every bubble bursts. Again.