Buttonwood – Beyond Reason
Stock markets are efficient machines, populated by rational investors seeking to make the best returns that they can. As evidence, look at the difficulty the average fund manager faces in trying to beat the market.
But if that is so, how do you explain the dotcom bubble, when companies with no profits and barely any sales had billion-dollar valuations? And what lies behind the continued existence of market anomalies, such as the tendency for smaller companies to outperform?
Academics have been discovering these effects for decades. There are seasonal patterns (stocks tend to do well in January and poorly during the summer). There are also valuation discrepancies (growth stocks tend to underperform). Some of these effects may be random. Analyze enough data and a few oddities will show up; plenty of people think some lottery numbers are "lucky" because they occur more often-though it would be odder still if they all turned up the same number of times. Other effects are real, but may be costly to exploit. For small companies, higher risk (smaller firms are more likely to go bust).
The underperformance of growth stocks is linked to an over enthusiasm for extrapolations. A company increases its profits at 20% a year for five years and investors are tempted to believe it can do so for 15; historically, however, such paragons are about as rare as vegetarian cats. In contrast, the prices of "value" stocks (which have a poorer record but lower ratings) perform better than expected.
The latest quirk to be examined is momentum, or the tendency for stocks that did well in the past to do well in the future. Three academics at the London Business School devised portfolios consisting of stocks that had outperformed over a 12 months period. They waited month to buy them and then rebalanced the portfolio after a further month. In the British market, the winners under this methodology outdid the worst performers by more than ten percentage points a year the past 108 years.
Some of this excess return will have been eaten up by higher costs; trading costs were substantial until the last 20 years. The momentum effect has also suffered some sharp switchbacks (including an 80% decline in 1975) that may have wiped out those traders who tried to follow it. That could be about to happen now, given that the strategy would leave investors heavily exposed to mining stocks. Momentum wannabes should also be warned that the effect works for individual stocks over months, not years.
Nevertheless, it is still a puzzle why such a glaring anomaly has not been arbitraged away. Academics have just about abandoned the idea that all investors are rational; there are too many examples of psychological quirks (such as an aversion to recognizing losses) for that to be the case. Nor can a bunch of "super-smart" investors necessarily keep prices in line; they may face constraints on their ability to trade or simply run out of money before the anomalies can be corrected.
Perhaps the most compelling reason why market prices are thought to beat is the "wisdom of crowds" phenomenon. If people are asked to estimate the number of jellybeans in a jar, their average estimate is usually quite close to the truth; indeed the average guess is far better than the vast majority of individual guesses. In other words, as Michael Mauboussin of the fund management group Legg Mason remarks, the collective is smarter than the average person within the collective.
But this wisdom depends on the diversity of the people making the guesses. Mr. Maubossin argues that problems occur when diversity breaks down and "group-think" starts to take over. Investors no longer guess how many jellybeans are in the jar, but what other people's guesses might turn out to be.
There's an old test that neatly makes this point. Participants have to choose a number between zero and 100 that will be two-thirds of the average choice of the others taking part. So if you thought the average would be 50, you would go for 33. However, if everyone makes this logical leap, the best guess should be 22 (two-thirds of 33). Extend this process a few times and you can work out that the best choice would be zero. In real life, however, not everyone is so rational and the correct answer is never that low.
In short, it is very hard to quantify the precise irrationality of investors. That is why investors get lured in to buying dotcom stocks in the hope that a "greater fool" will purchase them at a higher price. And that is why there will always be market anomalies for academics to discover.
No comments:
Post a Comment