Mean reversion strategy: Bet against it at your peril

Understanding mean reversion strategy

We’ve all heard and readily accept the investment axiom, buy low and sell high.  But have you ever thought about the market forces that makes this true?  Robert Shiller has. He won a Nobel Prize for this thinking.

In geek-speak that buy low/sell high market force is called mean reversion. The basic concept is that there is an average value for an asset class that roughly represents its fair value in the long run.  If the current value is well above that fair value, then the market tends to move that value back down towards that fair-value level. It’s a matter of when, not if. Conversely, if the current value of an asset class is significantly below that fair value, the market will tend, over time, to bring that value up.

Let’s say an imaginary Greek tech-sector index (as measured by the average price-to-earnings ratio of its constituent securities) has typically had a long-term fair value of 16.75. But right now, let’s pretend the current average market price of the Greek tech sector is 25. Mean reversion would tell us that eventually the value will move down from 25 back towards 16.75. Therefore, this would be the time to sell Greek tech stocks. Sell high, right?

The difference between knowing and following a strategy

Intuitively, most of us look at this phenomenon as obviously true, but the amount of academic firepower that has been aimed at proving this theory has been truly spectacular. And the number of investors who think they can second guess this axiom has been spectacular in a different, often more troubling way.

I see three primary reasons that make this is a complex phenomenon to predict:

• Equity markets are spectacularly and randomly volatile in the short-term.
• In the immortal words of Lord Keynes, “The market can remain irrational longer than you can remain solvent.”
• The market fueled by investor herding behavior inevitably overshoots, both on the upside and the downside.

I believe these three factors conspire to make it all but impossible to identify precisely when the reversion to the mean will commence.  However, it does not affect my faith that mean reversion exists. And if it exists, it should be understood and may be exploited.

Market inefficiency, Shiller and the Nobel Prize

Before Shiller, much of the work in forecasting stock prices focused on short-term time periods.  One of Shiller co-recipients in 2013, Eugene Fama, famously demonstrated in the 1960’s that markets are informationally efficient making forecasting of stock price movements in the short term effectively impossible to predict, using known information.  Shiller also won the Nobel Prize for Economic Sciences in 2013 for work that concluded stock price movements can be predicted over longer time periods (think five to seven years). Markets are therefore inefficient, in that, over time, they overshoot.

Shiller started with a basic understanding—that there is a tremendous amount of noise and randomness in market moves over shorter time periods, but looking to longer time periods can deliver insight into the likely path of stock values.

In his research, Shiller found that if a market’s current fair value was well below average, the longer-term future returns of the market were likely to be higher than the market’s long-term average return.

Stated differently, a below-average Cyclically Adjusted Price Earnings ratio (CAPE), seems to signal that the market will tend, over time, to bring that price up. And an above-average CAPE seems to signal that the market will tend to move that price back down towards that fair-value level.

So, if the long-term average CAPE of the imaginary Greek tech sector is 16.75 and the current CAPE is 25, Shiller says that eventually that prices (CAPE) will move down from 25 back towards 16.75.

In other words, over the span of several years, market prices tend to revert to the mean.

What this could mean for investors

Let’s get out of the imaginary and into reality. As of May 2, the CAPE of the U.S. equity market (S&P 500®) for April 2017 is 29.19¹. The long-term average CAPE since 1871 has been 16.75.

We believe Shiller. We don’t know when it will happen, but we fully expect that mean reversion will exert its inexorable force in the not-too-distant future.

As we stated in our Global Market Outlook – Q2 Update, the U.S. equity market has overshot and will likely underperform its historical return average over the next several years.  Therefore, we are underweight when it comes to U.S. equities.

When it comes to this particular equity market, it may take longer than we would like for the market to move in the direction we expect, but we are confident that it is just a question of when and not a question of if. And we believe Shiller would agree with us.

¹ Source: Yale Shiller Data. http://www.econ.yale.edu/~shiller/data/ie_data.xls . April 2017.