Friday, February 13, 2009

"I think it's the duty of the comedian to find out where the line is drawn and cross it deliberately." - George Carlin

***

Now is the time for a revolution in economic thought

"Consider the following passage:

“Most economic theorists have been going down the wrong track. When economic models fail, they are seldom thrown away. Rather they are ‘fixed' - amended, qualified, particularised, expanded and complicated.

“Bit by bit, from a bad seed a big but sickly tree is built with glue, nails, screws and scaffolding. Conventional economics assumes the financial system is a linear, continuous, rational machine and these false assumptions are built into the risk models used by many of the world's banks. As a result, the odds of financial ruin in a free global market economy have been grossly underestimated. By using such methods there is no limit to how bad a bank's losses can get. Its own bankruptcy is the least of the worries; it will default on its obligations to other banks - and so the losses will spread from one inter-linked financial house to another. Only forceful action by regulators to put a firewall round the sickest firms will stop the crisis spreading. But bad news tends to come in flocks and a bank that weathers one crisis may not survive a second or a third.”

This uncannily precise description of the present crisis above was not written by an economist... Modern economists were inherently incapable of understanding such a problem because they assumed that investors were “rational” and markets “efficient”.

These assumptions led inevitably to disaster once they were blown apart. The author who came so close to understanding the true causes of the present crisis was not an economist but a mathematician.

Benoît Mandelbrot, a towering figure of 20th-century science, who invented fractal geometry and pioneered the mathematical analysis of chaos and complex systems...

At the other end of the academic spectrum, we find economists treating ideas from sociology, psychology or philosophy with the same derision and disdain. George Soros is no mathematician like Mandelbrot, but he has repeatedly demonstrated far better understanding of how market economies work than any professional economist by using psychological and philosophical ideas. His books have explained convincingly how false beliefs among investors can create self-reinforcing boom-bust cycles of exactly the kind afflicting the world today. But the reaction to these ideas has been the same as to Mandelbrot's: a complacent refusal among academic economists, regulators and central bankers even to think seriously about approaches that challenge the central orthodoxies of modern economics: that “efficient” markets inhabited by “rational” investors send price signals which, in some sense, are always right...

A control theory approach, used by serious mathematicians such as Nicos Christofides and Shahid Chaudhri, working at the Centre for Quantitative Finance at London's Imperial College, has applied advanced mathematics from aerodynamics and control engineering to analyse financial turbulence without the over-simplified assumptions, such as continuous liquidity, which have caused the recent disasters in risk management and regulation.

But the challenge that existing economic orthodoxy may find most disconcerting is Imperfect Knowledge Economics (IKE), the name of a path-breaking recent book by Roman Frydman and Michael Goldberg, two American economists. Building on ideas of Edmund Phelps, one of the few Nobel Laureate economists who rejected the consensus view on rational expectations, IKE uses similar tools to conventional economics to generate radically different results. It insists that the future is inherently unknowable and therefore that there is always a multitude of plausible models of the way the economy works.

With this obvious, but critically important, change in assumptions, IKE demolishes most of the conclusions of rational expectations. More importantly, it shows that reasonable assumptions about economic uncertainty can produce financial models that give less spurious accuracy than the rational expectations models but are statistically far closer to what happens in the real world."
blog comments powered by Disqus