[chalst]

The case that subprime mortgages were in a bubble was made very efficiently in public by several people - I was subscribed to Nouriel Roubini's widely followed blog at the time; Gillian Tett at the FT also did good work exposing what was going on with property prices - and believed by many people. The bubble continued for years after it had convincingly been called.

However, relatively few people figured out how to efficiently make use of this information. Shorting the stocks of banks like AIG is very high risk, and can cost you everything if you don't know when the bubble is going to burst. In the end, the only shorting strategy that was effective was to use CDS, a newly popular derivative (which was a very smart idea). Jon Paulson's hedge fund made $15 billion from this, but Kyle Bass, another person who made money from this, says only 15 people figured out the CDS trade and made money from it. The amount of shorting that happened was tiny in proportion to the size of the bubble.

The EMH is a hypothesis that was contrived in an analysis that does not take account the actual way that information gets incorporated into prices, the actual trades available to market participants, or the asymmetry between upside and downside risks (aka. the Keynesian risks of opposing the market's animal spirits).

The EMH states that market prices correctly price in all public information. The weaker proposition, that it is hard to get rich quick from spotting market mispricing using just public information, does seem to be true.