Mean reversion is the one thing we can, eventually, rely on to align price with value. It also runs counter to the collective thinking in markets.
Mean reversion is the idea that things – growth rates, earnings, prices, returns – eventually move toward an average. It’s what you get when short-term expectations conflict with long-term reality.
History is littered with companies once considered great, that have since fallen by the wayside. And the companies that were written off for dead often exceed their lowly expectations by briefly performing well.
It turns out, few companies are immune to the cycle of creative destruction (it’s a feature, not a bug, of capitalism). Mean reversion is what results. Tobias Carlisle explains why in his book Deep Value:
Mauboussin’s research supports Graham’s view that, while some businesses do generate persistently high or low returns on invested capital beyond what chance dictates, there exists a strong tendency toward mean reversion in most businesses. For those businesses that do post sustainable high returns, Mauboussin could not prospectively identify the factors behind those sustainable high returns. In other words, they cannot predict which businesses will sustain high returns on capital and which will revert back to the mean, which is the average experience. Recall that Graham warned about constituting notions of “a good business” on the basis of a superior record:
Abnormally good or abnormally bad conditions do not last forever. This is true of general business but of particular industries as well. Corrective forces are usually set in motion which tend to restore profits where they have disappeared, or to reduce them where they are excessive in relation to capital.
Mauboussin observes that reversion to the mean is a powerful force, and it impacts return on invested capital as it does many other data series.
…
Here is the simple truth: mean reversion is pervasive, and it works on financial results as it does on stock prices. On average, super-normal returns on capital revert to the mean. Only special cases avoid mean reversion, and the factors that separate the also-rans from the special cases are impossible to identify prospectively.
Rarely do companies have an impenetrable moat – an advantage that protects it from all competitors. So great companies become less great. Bad companies become less bad. And stocks prices reflect that reality…eventually.
But mean reversion isn’t only driven by competitive forces. Investor behavior has a role too.
Mean reversion in the markets looks a lot more like the gambler’s fallacy made real — movements in security prices, individually and in aggregate, tend to be followed by subsequent price movements in the opposite direction. The more extreme the initial price movement, the greater will be the subsequent adjustment in the opposite direction. This phenomenon is observable in the financial results of businesses, in the market prices of securities, in the movements of indices, and in the results of investment managers. The reasons are manifold, but the most obvious is that the trials aren’t independent — our own trading decisions are affected by the buying or selling preceding our trade. English economist John Maynard Keynes observed in The General Theory of Employment, Interest, And Money that “[d]ay-to-day fluctuations in the profits of existing investments, which are obviously of an ephemeral and nonsignificant character, tend to have an altogether excessive, and even an absurd, influence on the market.” Two economists known for research into both market behavior and individual decision-making, Werner De Bondt and Richard Thaler, theorized that it is this overreaction to meaningless price movements that creates the conditions for mean reversion. De Bondt and Thaler speculated that investors overreact to short-term, random fluctuations in market prices, and this overreaction causes the stock prices to temporarily depart from intrinsic value.
Investors overreact to good news and bad news. They overestimate the impact it will have on a company. They become too optimistic or pessimistic. And the overreaction gets reflected in stock prices.
But the study, Carlisle refers to, showed unexpected results.
De Bondt and Thaler’s findings stand the conventional wisdom on its head and show compelling evidence for mean reversion in stocks in a variety of forms. Big gains and drops don’t persist. Rather, extreme drops in stock prices tend to be followed by big gains and big drops follow extreme gains. Big drops in stock prices also tend to be followed by significant earnings increases and significant stock price increases are followed by slower rates of increase or declines in earnings.
The losers outperform the winners. Why?
Howard Marks would sum it up as the result of first level thinking. It’s a great company, so it’s a buy. It’s a poor company, so it’s a sell.
Investors fail to account for the possibility of mean reversion.
They set high expectation based on good news. High expectations set high prices. High expectations not only have to be met but exceeded for prices to move higher.
Since high prices leave little margin for error, one slip will send prices in the opposite direction. Simply, investors set an increasingly higher hurdle that is harder to clear. Greatness gets priced in.
And inversely, it happens with stocks given too low expectations too. Low expectations set low prices. When expectations are too low or nonexistent, the slightest improvement can set prices soaring.
The counterintuitive result is that value ends up beating glamour – the worst performers end up outperforming the best performers over time.
Because future growth is hard to predict.
Because high growth companies eventually become slow growth companies.
Because competition is relentless.
Because investor overreaction drives prices too far in both directions.
Because the objective is not to pick existing winners, it’s to pick mispriced stocks.
Because of mean reversion.
Source:
Deep Value by Tobias Carlisle
Related Reading:
Picking the Worst Performing Countries
Putting the CAPE Ratio Into Context