Global Agenda: Black box blues

In contemporary Western society, markets are the nearest thing to organized religion.

global agenda 88 (photo credit: )
global agenda 88
(photo credit: )
JOAN (to Canon de Courcelles, a young priest of 30): Thou art a rare noodle, Master. Do what was done last time is thy rule, eh? COURCELLES: Thou wanton: dost thou dare call me noodle? From Bernard Shaw's Saint Joan, Scene vi. 'Twas ever thus. The pompous representatives of the Establishment cannot bear to be exposed in public as "noodles," locked into a narrow mindset based on "doing what was done last time." In the secular, post-modern age, priests and religious functionaries hardly count as the Establishment anymore, and certainly not as its foremost representatives. In contemporary Western society, markets are the nearest thing to organized religion and those people with proven ability to make money in the markets are the priesthood, venerated by the ignorant masses. What has changed, noticeably, is that age is no longer a requisite factor - the socio-economic ladder can be scaled in fewer years than pre-moderns would have believed possible. This has been made possible by technological advances that enable younger people to rise very quickly - indeed, they give youth a strong advantage over age. There remains, however, the inbuilt advantage that age carries with it, namely experience. What can youth do to neutralize the disadvantage stemming from lack of experience? The natural response of today's younger generation is to use technology - its own inherent advantage - to acquire the experience that the oldsters have and it doesn't. The result is the rise of "black box investing," known more politely as "quantitative analysis," or simply "quant." In quantitative analysis, you start by amassing historical data on whichever security or other financial instrument, or market index, or whatever, you are interested in. You then use this data to analyze how the said share/currency/index behaved over time, and in relation to any or all of a host of other instruments or economic and financial data. Did this share, or this kind of bond, tend to rise or fall more than other similar instruments when, say, interest rates fell, or during recessions? This analysis ultimately generates patterns which can be mapped and then compared and contrasted with other instruments. A simple example would be that the Australian dollar rises much more than the Swiss franc in periods when commodity prices rise. The practical conclusion would be that once you can identify a clear upward trend in commodities, the Aussie is a better investment than the Swissie (and probably that the Sydney stock exchange is preferable to the one in Zurich). But you can go further: instead of merely taking your money and putting it in Australia rather than Switzerland, you can borrow other people's money - Swiss money, of course - and add it to your own. You have now constructed a 'model' and you have used it to "leverage" your "position" (i.e. bet). All that remains for you to do is to hire a Phd in math and/or computer science, make this analysis much more complex and then create equations that summarize the model and generate investment instructions, and put the whole package into a computer. You now have a black box. Next, using your historical data, show how it would have yielded hefty profits over the years - and proceed to launch a hedge fund, in which individuals and institutions with more money than ideas of what to do with it can invest. You take 2 percent per annum of the money invested as a management fee and 20% of the profits as a performance fee. In normal circumstances, you can't go wrong and your future is assured. However, in abnormal circumstances, the model malfunctions and the black box seizes up, causing your fund to make heavy losses. How abnormal do these circumstances have to be? According to the quant analysts, the chances of them happening range from once in 10,000 to once in 10 million years, depending on what parameters the model uses. In practice, however, successive generations of black boxes have blown up in 1987, 1994, 1998 and 2007 - with some significant hiccups in between. Why this should be is already the subject of frenetic research across a range of disciplines. But the answer has to be sought among humans, not machines. Because the Establishment is inherently hostile to people who think outside the box, it is populated by noodles who can only build machines programmed to do what was done last time, not to anticipate what might happen next time.