This is part 2 in my series on why and how the 2008 financial crisis is relevant to economics. The first instalment discussed why the good times during the boom are no excuse for the bad times during the bust. This instalment discusses the use of the Efficient Markets Hypothesis (EMH) to defend economists’ inability to forecast the movements of financial markets, hereafter referred to as the ‘EMH-twist’.
Argument #2: ““The EMH claims that crises are unpredictable, so the fact that economists didn’t predict the crisis is not a problem for economics at all.”
As far as I’m aware, this argument was first used by John Cochrane, and it has reappeared multiple times since then: for example, it was more recently referenced by Andrew Lilco, who was sadly echoed by the generally infallible Chris Dillow. The idea is that financial markets process new information faster than any one individual, government or institution could, and so for most people they may seem to behave unpredictably. However, economists can not be expected to understand these sudden movements better than anyone else, so expecting them to foresee market crashes is absurd. As Cochrane puts it, “it makes no sense whatsoever to try to discredit efﬁcient market theory in ﬁnance because its followers didn’t see the crash coming”.
However, this logic is completely circular. The mere fact that a theory exists which claims crises are unpredictable does not mean that, if a crisis is not predicted – particularly by the proponents of said theory – this shows the theory is correct. If the EMH had, to the best of our knowledge, been shown to be correct, then the EMH-twist might hold some water, but we must establish this truth separately from the the fact its proponents didn’t predict the crisis (David Glasner recently made a similar point about the ubiquitous use of rational expectations in macroeconomics). While Cochrane does claim that the central tenet of the EMH “is probably the best-tested proposition in all the social sciences”, he fails to reference supporting evidence, and in fact goes on to add substantial qualifications to the empirical record of the EMH, admitting that market volatility might happen “because people are prey to bursts of irrational optimism and pessimism”.
It is not necessarily my aim to establish the truth or falsity of the EMH here: it has been discussed extensively elsewhere. However, there are a couple of key tests for whether or not it applies to 2008. The first is whether or not anybody – both adherents and detractors of the theory – foresaw the crisis. While the EMH claims nobody could, this is clearly wrong: some people in finance made a lot of money; some economists not only called it but had frameworks that explained it well once it happened; quite a few people (even mainstream economists) at least noted the existence of a housing bubble. The EMH can attribute these predictions to simple luck, but now we’re back to circularity: assume the EMH is true, then appeal to it to rationalise any possible market movement. The second test of the EMH, since it depends on new information to trigger volatility, is to ask exactly what new information became available just before the crash. However, the financial instruments key to 2008 were used by investment banks for a good few years prior to the crash, so it’s quite difficult to claim that new information about these suddenly became available in 2007-8. Instead, what happened was a collective realisation that everyone knew very little about the products they’d been trading, resulting in a classic panic.
In fairness, there is an element of truth to the EMH-twist. Financial markets are incredibly difficult to understand, and the argument that economists don’t yet understand them, along with a mea culpa, might be acceptable – there are many things natural scientists still don’t understand, such as dark matter, or what happened ‘before’ the big bang. However, the EMH-twist as used by Cochrane et al is phrased more strongly: it is the assertion that economists can’t and shouldn’t understand the movements of financial markets, simply because the EMH allows them to wash their hands of the task. We wouldn’t accept this kind of attitude from any other field, so I can’t help but feel Cochrane’s claim that “the economist’s job is not to ‘explain’ market ﬂuctuations after the fact” can only be met with: “then what is the economists’ job, exactly?”
The next instalment in the series will be part 3: econoracles.