The third anniversary of the peak of the NASDAQ boom has produced a fair bit of commentary. As an economist, I think the biggest impact, still only dimly-perceived, is the refutation of the efficient markets hypothesis.

I’ll start with the introduction from a piece I wrote for the Fin last year, which ran under the titleMarket theory unravels

Although the name gives the general idea, the efficient markets hypothesis is hard to state in simple terms. To complicate matters further, the hypothesis comes in a variety of intensities, from weak to very strong. The weakest version simply says that it is impossible to predict shares prices based on their past behavior. This is bad news for day-traders and chartists, but does not matter much to the rest of us.

In the strongest version of the efficient markets hypothesis, market prices for assets such as shares represent the best possible estimate of their value, taking account of all available information, public or private. Moreover, markets yield the most efficient possible allocation of risk. Markets are not perfect, but, they are claimed to be better than any alternative institution, including governments.

Although the statistical evidence generally doesn’t support strong versions of the efficient markets hypothesis, statistical testing is never going to finally refute an idea that’s so obviously appealing to a lot of people.

But I think the experience of the NASDAQ bubble has fatally undermined the efficient markets hypothesis, for several reasons.

First, in terms of scale, financial magnitude and worldwide publicity, it was far bigger than any previous bubble. Trillions of dollars in market capitalisation were lost and at least one trillion dollars was dissipated in investments that ranged from unsound to absurd. Millions of individual investors made and lost large amounts of money.

Second, it was widely recognised and denounced as a bubble, even as it developed. Here’s a piece I wrote in 1999 criticising dotcoms in particular, but the speculative nature of the boom had been recognised as early as 1996, when Robert Shiller coined, and Alan Greenspan popularised the phrase ‘irrational exuberance’.

Third, all the ‘checks and balances’ that are supposed to prevent a bubble were present and all were ineffectual or counterproductive. It’s hard to recall, but there was a time when people believed that stock market analysts served a positive role in informing financial markets. Accountants, debt rating agences and regulators have been as thoroughly discredited as analysts, but all clearly failed to their job.

Most notable of all, but less noticed was the failure of counter-speculators to burst the bubble. George Soros lost billions in 1998 and 1999 by short-selling stocks that were grossly overvalued. By the time the bubble reached its peak three years ago, the bears had all retired to lick their wounds – it was only when the bulls finally ran out of money that the madness stopped.

The implications of rejecting the efficient markets hypothesis are profound. As I observed in the AFR piece quoted above

Judging by the experience of other bubbles, share markets will take years to recover from the 1990s. The social and political implications will emerge even more slowly. Political programs like privatisation, cultural attitudes based on the idolisation of ruthless CEOs, and much of the ideological framework of the last two decades have been founded on the efficient markets hypothesis. They will not vanish overnight. But they are doomed, nonetheless.