Because I was a dissatisfied economics major in college (it's a day of remniscing around this blog), I took a class with what turned out to be the world's leading Newton scholar, George Smith, who also happens to be an expert on aerodynamically induced vibration and resulting metal fatigue failures in jet engines and other turbomachinery (very complex systems; George probably dreams about the Navier-Stokes equation). George always emphasized that in science one can learn from failure (and one of his great insights is that Newton taught us how to do this in a systematic and evidentially interesting fashion). And so the title of this week's blog is a homage to George, who always encouraged my interest in the philosophy of social science. Next week I'll return to more conceptual questions at the heart of modern mathematical finance theory, but this week I'll muse a bit more on the market as a complex system.
Last week (on a tip by technology guru Michael Krigsman) I called attention to a lovely, short piece "How Complex Systems Fail" by Richard Cook, that is 18 (the Kabbalist in me rejoices!) very smart bullet points on the nature of failure of complex systems. In the discussion, Krigsman also called attention to this piece in the Financial Times, which makes same useful comparison between markets and nuclear power stations as complex systems (although its rhetoric makes it -- misleadingly -- sound as if the folk that were marketing collateralised debt obligations (CDOs) had no idea they were promoting junk for profit).
By ethos I don't mean that market participants should forego the profit motive and somehow become moral or follow a code of ethics. Rather, I am thinking in terms of Enlightened self interest (not to mention that there is a moral imperative for all of us to think about the matter: the cost of collapse are often born worst by relatively innocent bystanders [folk like you and me]).
How can one create an ethos? Here Cook is useful: by training and practice. That is, (nearly) all complex systems (with intentional agents in it) that are successfully defended from collapse are ones in which major participants prepare for crises regularly. They simulate these and they train for them. (Think of soldiers, surgeons and their staff, firemen, nuclear power station operators, EMTs, etc.) This promotes -- as Cook emphasizes -- a constantly evolving know how about the vulnerabilities of the system.
So, I propose that one Friday each month, after the market closes, there are major real time simulations of impending collapses. (Let's say for two hours--this may hurt turn-over at a few bars in London, New York, Hong Kong, etc [I will be shorting Heineken, and will buy calls on Dominoes Pizza]. But it is also realistic because most decisions in a crisis have to be taken after markets closed during a weekend.) This way the major participants (which are often huge, complex bureaucracies) can learn to prepare for crises. Moreover, CEOs/COOs and outsiders (regulators, central bankers, treasure departments, etc) can learn to ask the right questions and start to map the evolving system and its vulnerable nodes (in the language of network theory). And folk will learn to look at their own organization and relationships with others with fresh eyes rather than responding largely to the last crisis. So, such simulations are training and learning opportunities and in doing so regularly one builds an ethos (which is really just to say, that one has good habits), in which major market participants also become defenders against collapse. That is, such an ethos needs to become a dynamic one and that accepts in the words of Cook, that "catastrophe is always around the corner." (I propose that the Federal Reserve hires a few philosophers of economics <grin> to work as consultants on preparing and guiding the simulations!)
A new ethos is, of course, not the whole solution. I also favor more limited, technical approaches: 1. unlimited liability for the partners of financial institutions; 2. the imposition of a tobin tax so as to discourage trading on on the order of milliseconds (that you Jared Woodard for discussion); and a few others another time.