Methods for all moments

THE “prize in economic sciences in memory of Alfred Nobel”, as it is officially known, sometimes struggles to command the same respect as its counterparts. Though awarded by the Royal Swedish Academy of Sciences, just like the prizes in physics, chemistry and medicine, it was a latecomer to the ceremonies, established in 1968 by Sweden’s central bank rather than in 1896 by Mr Nobel’s will. This year’s winners appeared to reinforce doubts about the prize’s standing. One, Eugene Fama of Chicago, is known for his ardent belief in the efficiency of markets: he declined to renew his subscription to this newspaper after tiring of its incessant warning about bubbles, the very existence of which he denies. Robert Shiller from Yale, in contrast, is known for his prescient warnings of bubbles, in technology stocks in the 1990s and in housing in the 2000s.

Yet for all the apparent contradiction, Messrs Fama and Shiller, along with Lars Peter Hansen (also of Chicago), have established most of the surprisingly small amount economics has to say about asset prices. Although they disagree with one another about how markets operate, the work for which they are being recognised is not itself irreconcilable. For proof, look no further than investors, who have profited from the work of all three men.

Mr Fama began studying data on asset prices while working on his doctoral dissertation in the early 1960s. He set out to test the hypothesis that markets are efficient—that stock prices incorporate all available information immediately and are therefore entirely unpredictable over the short term. His analysis revealed that this is true over horizons of days or weeks, or at least that markets are efficient enough that no trader could consistently profit from a stock price’s predictability after taking into account transaction costs. With co-authors Mr Fama pioneered the use of the “event study” in the analysis of asset prices. By tracking prices in the days immediately before and after important news breaks, they demonstrated that markets react quickly and then once again become unpredictable. An influential article entitled “Efficient Capital Markets”, published in the Journal of Finance in 1970, laid out a research programme for asset-price analysis. Subsequent work has provided a wealth of evidence underlining the efficiency of markets over short periods.

Mr Shiller entered the debate in 1981 with a paper in the American Economic Review entitled “Do Stock Prices Move Too Much to be Justified by Subsequent Changes in Dividends?” It concluded that they did. Although efficient-market theory would suggest that share prices should simply reflect the latest information on the value of the revenues they will yield, they are in fact far more volatile than dividends. Mr Shiller’s work implied that, given some knowledge of the underlying trend in the price of a stock, one could predict future movements. He devised a modified price-to-earnings ratio for shares which used a rolling ten-year average of earnings instead of current earnings. He reckoned stocks that looked overvalued by that measure could be expected to fall over time: an insight that guided his warnings of bubbles in both the 1990s and 2000s. It was after a conversation with Mr Shiller that Alan Greenspan, then head of the Federal Reserve, said high share prices might reflect “irrational exuberance”.

Whereas Mr Fama’s work demonstrated that short-run prices were unpredictable, Mr Shiller’s showed that over periods of several years assets could move in foreseeable ways. Mr Fama does not disagree entirely: one of his papers found that short-run interest rates often influence longer-term trends in the stockmarket. Research he did with Kenneth French of Dartmouth College indicated that the longer the period one considered, the more predictable stock returns became. But he still disputes the idea that asset prices can lose all purchase on the information at hand.

Mr Shiller, meanwhile, sees a role played by individual and crowd psychology, drawing on the work of Daniel Kahneman, a psychologist turned economist who won the Nobel prize in 2002. (Indeed, the Swedish academy might plausibly have paired Mr Shiller with Richard Thaler, another academic at the University of Chicago, in a Nobel prize for behavioural economics.) Other economists reject the behavioural approach, maintaining that predictable returns reflect investors’ rational insistence on compensation for holding riskier assets.

In 1982 Mr Hansen developed what has since become a very influential statistical technique known as “generalised method of moments estimation”. This helps econometricians test theories and make the best use of what information they have (even when it is not all they would like to have), by identifying which of various potential explanatory variables are most likely to deliver statistically robust results. Since then, much of his work has focused on understanding the linkages over time between the prices of assets and macroeconomic variables such as total consumption—a fast-growing field known as macro-financial modelling. Among the questions he has looked at is how the business cycle influences asset prices, and vice versa. His models typically involve people having to make decisions without all the information they would like and with considerable uncertainty about the future, sometimes resulting in asset-price behaviour that seems relatively efficient and at other times quite irrational.

Back to work

What all three laureates share is a commitment to backing up theoretical work with rigorous empirical analysis. The world of finance could do with more of that sort of thing, judging by the wild mispricing of assets revealed by the crisis. Shortly after being told of his award, Mr Hansen was asked how well economists are doing in understanding asset prices. “We are making a little bit of progress,” he replied, “but there’s a lot more to be done.”

_{Economist.com/blogs/freeexchange}

Explore our A to Z of economics