On efficient markets, bubbles and monkeys

In one of the most bizarre pieces of reasoning of all time Eugene Fama has told John Cassidy in the New Yorker that the GFC was due to a global recession which meant people couldn't pay their mortgages. Most of us thought the recession followed this, but not acording to Fama.

You see in Fama's world rational bankers would nly have lent money to people who could repay them, so for people not to be able to repay there must be some exogenous cause, namely a recession. Thankfully Fama admits he is not a macroeconomist - because no macroeconomist could find evidence of an ACTUAL recession before housing prices started to decline.

Fama is relying on a perfectly reasonable - though perverse - type of reasoning. I assume X is rational, therefore everything X does is rational. He simply doesn't allow for the three facts of the GFC - (1) herd behaviour in pricing housing assets (see below), (2) a flawed risk model that assumed the risk of simultaneous downward movement of prices in housing couldn't happen and (3) herd behaviour in pricing credit assets (see below).

Ultimately Fama's defence of his efficient markets hypothesis is that at all times the market accurately reflected all the known information, it just couldn't factor in the unknown information which was about a recession...or what he called "economic activity" which economists can't help us with.

Is it any wonder that Lusha the monkey has reportedly outperformed 94% of Russian bankers? Of course, this is finance not economics. But I guess they all have the perfect Fama excuse. "We accurately predicted the market, it is just that the world changed in unexpected ways".

The whole Fama interview is worth a read.

Meanwhile over at The Economist last week they were again warning about asset bubbles. Most notably they said "In housing, a measure based on rents shows that American prices are back to fair value but prices in Britain, France, Spain and Australia are all 30-50% above their historic averages. Low mortgage rates (and government schemes to head off foreclosures) have stopped prices falling to the lows of previous downturns."

This does raise the really interesting question of prices - and whether asset prices are best measured on fundamentals (the total return you think you could generate by USING the asset) or by markets (the return you could generate by SELLING the asset). The problems emerge over what to do when the two diverge - especially what to do from an accounting point of view.

Accounting is not counting - it is a artificial construct designed to provide useable information for decision making. In fact accounting is a very Feyerabendian science. A key issue in the GFC was "mark to market" as an accounting rule. It was introduced as an accounting rule to stop firms inflating their valuations by dreaming up future revenues - masthead revaluations in newspapers were classics.

But it had disastrous consequences in the GFC. Firstly banks lookd to be making a fortune by marking tomarket all sorts of assets fuelled by an asset price bubble. But worse was the need to mark to market as they came down.

In reality a far better view of accounting would be assymetric. Assets should be valued only on fundamentals if you want to revalue them up, but must be marked to market if market prices are below the valuation on fundamentals.