Tight Coupling and Complexity
From The London Review of Books comes an article on the rise of algorithmic trading:
Systems that are both tightly coupled and highly complex, Perrow argues in Normal Accidents, are inherently dangerous. Crudely put, high complexity in a system means that if something goes wrong it takes time to work out what has happened and to act appropriately. Tight coupling means that one doesn’t have that time. Moreover, he suggests, a tightly coupled system needs centralised management, but a highly complex system can’t be managed effectively in a centralised way because we simply don’t understand it well enough; therefore its organisation must be decentralised. Systems that combine tight coupling with high complexity are an organisational contradiction, Perrow argues: they are ‘a kind of Push-me-pul-lyou out of the Doctor Dolittle stories (a beast with heads at both ends that wanted to go in both directions at once).
Perrow’s theory is just that, a theory. It has never been tested very systematically, and certainly never proved conclusively, but it points us in a necessary direction. When thinking about automated trading, it’s easy to focus too narrowly, either pointing complacently to its undoubted benefits or invoking a sometimes exaggerated fear of out of control computers. Instead, we have to think about financial systems as a whole, desperately hard though that kind of thinking may be. The credit system that failed so spectacularly in 2007-8 is slowly recovering, but governments have not dealt with the systemic flaws that led to the crisis, such as the combination of banks that are too big to be allowed to fail and ‘shadow banks’ (institutions that perform bank-like functions but aren’t banks) that are regulated too weakly. Share trading is another such system: it is less tightly interconnected in Europe than in the United States, but it is drifting in that direction here as well. There has been no full-blown stock-market crisis since October 1987: last May’s events were not on that scale.[*] But as yet we have done little to ensure that there won’t be another.
I highly reccommend reading Normal Accidents and The Logic of Failure: Recognizing And Avoiding Error In Complex Situations.