Except in economics. Economic hypothesis lead a charmed life: Counterexamples don’t kill them and their popularity mostly depends on how closely they resemble some ivory tower notion of how the world should work.
There are many economic concepts that are manifestly untrue, but the Efficient Market Hypothesis (EMH) is probably the most entrenched and most blatantly wrong idea out there. There are several versions of EMH, but they all posit that prices of traded assets (stocks, bonds, etc.) reflect all available information.
Eugene Fama came up with the idea in the ’60. It was certainly a good hypothesis because it is plausible, it makes specific predictions, and it is testable. In this post we will list a few of the more colorful counterexamples that should have killed this hypothesis well before it reached voting age.
|S&P 500 ’87 Crash|
The most dramatic counterexample is the crash of ’87. On October 19th 1987 the S&P 500 declined 20.5%. According to EMH, this has to reflect information about the underlying economic situation that became known on that day. This change has to be big enough to warrant 20% lower valuations, yet unnoticeable enough that nobody suspected anything before the 19th.
You can search the news archives, and you won’t find anything that fits the bill. In fact, the only think you will find, is news of the market crash. So if no material new information became available, this market move cannot be explained by EMH. The only explanation is that markets are not efficient. That should have been the end of the story, but even 24 years later, Wikipedia has a long entry on EMH. That is extraordinary longevity for a failed idea!
|Intraday DOW values for 5/6/2010 (from Wiki)|
An EMH defender may try to argue that this is a single exceptional day and doesn’t really count, so let’s look at another example.
Last year on May 6th the Flash Crash happened. This crash and the subsequent recovery played out over a the span of a few hours. The Dow Industrial index lost around 9% only to gain most of it back an hour later.
This one is even harder to explain with HMH because the price moves imply that some terrible news hit the market, which was offset by some equally dramatic good news moments later. Needless to say, a news scan does not reveal anything that could account for this.
Either one of these examples should have been enough to consign EMH to the junkyard of wrecked ideas, but one could still argue that both of these are sharp shocks to the financial system, and that maybe EMH fails during crises. Of course that means that EMH is completely useless because the only time one needs financial modeling is when a crisis hits. Nevertheless, let’s take a look at a market inefficiency that proceed at a slower pace.
According to EMH it is impossible to beat the market. Generating above average returns would require some kind of an information advantage that allows the investor to pick winners. EMH posits that all available information is already reflected in prices, so it is impossible to have information that is not priced in and therefore impossible to consistently outperform the market.
|Momentum Bets (from Economist)|
For the last 30 years it has been well known that it is in fact very simple to outperform the market: Simply buy the stocks with the biggest price increases over the last year, hold them for a year, and repeat.
The Economist has a very nice piece on this in the January 8th issue. The chart at right (from the Economist) shows the cumulative returns from buying the 20% top performing stocks (light blue) vs. the middle 60% and the bottom 20% over the last 110 years.
In an efficient market the three lines would be on top of each other. How much a given stock has gained in the past 12 months is public information. The implications for the future performance should already be priced in. Specifically, people should be bidding up 20% best performing stocks so much, that their future gains drop to the market average. The chart clearly shows that this is not the case.
Note that the scale is logarithmic. Investing 1 British Pound into the bottom 20% turns it into 49 Pounds in 2009. Investing it into the top 20% turns it into 2.3 Million Pounds. This is an example of a dramatic and persistent market inefficiency that cannot be chalked up to special circumstances.
EMH can’t claim to have anything to do with reality, but apparently it is a sufficiently useful myth that it is still going strong today.