10th November 2010
"Yes, this model is deeply flawed but it's important to recognise that most economists understand this.
"The problem is that to replace it you need a model of how human behaviour impacts markets. We don't have that model and it's entirely possible that we may never do so.
Behavioural economics, at least at the macroeconomic level, is really an extension of the old efficient market hypothesis models. This relationship is hidden behind the rhetoric, but simply stating that market gyrations are somehow down to "behavioural biases" merely replaces one set of mystical incantations with another.
"It sounds plausible, but once upon a time so did volcano gods.
"What's more interesting is that people are rediscovering that the world is an uncertain place and that we can't, ever, predict everything that will happen. Frank Knight pointed that out 80 years ago, but the lessons rather got lost along the way.
"When economists start building models like engineers rather than like physicists we'll all be better off.
"There are some hopeful signs, but there are a lot of people with heavy emotional baggage in the old models: behavioural biases affect economists, too."
"There's some evidence that contagion can spread through networks like the internet but market truths are eternal, not built on technology: 1929 was a global crash.
"The usual transmission factor for these panics is liquidity, that's been true for hundreds of years, and most liquidity is in the hands of institutions not Tweeters (although that may not be true for very small cap stocks).
"The recent increase in asset class correlation is most likely a temporary phenomenon caused by the realisation that when there's a liquidity crisis all assets not backed by a lender of last resort are at risk.
"This will pass with time and the fading of memory: that's disaster myopia for you.