"The EMH has two components that I call "The Price is Right" and "No Free Lunch". The price is right principle says asset prices will, to use Mr Fama’s words "fully reflect" available information, and thus "provide accurate signals for resource allocation". The no free lunch principle is that market prices are impossible to predict and so it is hard for any investor to beat the market after taking risk into account."
And on the "price is right" component and examples of the deviation from it,
"As early as 1984 Robert Shiller, the economist, correctly and boldly called this 'one of the most remarkable errors in the history of economic thought'. The reason this is an error is that prices can be unpredictable and still wrong; the difference between the random walk fluctuations of correct asset prices and the unpredictable wanderings of a drunk are not discernable... The prices of closed-end mutual funds (whose funds are traded on stock exchanges rather than redeemed for cash) are often different from the value of the shares they own. This violates the basic building block of finance – the law of one price – and does not depend on any pricing model... When 3Com, the technology company, spun off its Palm unit, only 5 per cent of the Palm shares were sold; the rest went to 3Com shareholders. Each shareholder got 1.5 shares of Palm. It does not take an economist to see that in a rational world the price of 3Com would have to be greater than 1.5 times the share of Palm, but for months this simple bit of arithmetic was violated."
While clearly stating that "markets can be wrong and the price is not always right", he draws two lessons for the "free lunch" component,
"The first is that many investments have risks that are more correlated than they appear. The second is that high returns based on high leverage may be a mirage. One would think rational investors would have learnt this from the fall of Long Term Capital Management, when both problems were evident, but the lure of seemingly high returns is hard to resist. On the price is right, if we include the earlier bubble in Japanese real estate, we have now had three enormous price distortions in recent memory. They led to misallocations of resources measured in the trillions and, in the latest bubble, a global credit meltdown. If asset prices could be relied upon to always be "right", then these bubbles would not occur. But they have, so what are we to do?"