Smoothed O"PE"rator

PE. To schoolchildren, it’s the (typically) most fun hour of the day. To investors with a hankering for numbers, though, it’s a popular valuation metric: the price-to-earnings ratio (P/E). Like many valuation tools, P/E stars in a host of old wives’ tales—a common one being the assumption markets tend to produce sub-average future returns when P/Es are high, and vice versa. Some investors believe following this logic is key to outperforming over time. To me this seems fraught with peril—P/Es can be useful, mostly as a way to illustrate sentiment, but they don’t perfectly forecast market trends. They rely on backward-looking information and past performance, which never predicts future returns.

Still, many investors believe in P/Es’ predictive abilities. They see high valuations as a cue to sell and low valuations as a buy signal, essentially using P/Es as a timing tool. Problem is, this assumes past performance is predictive—it isn’t. It also assumes price changes drive valuations while the denominator—earnings—is fixed. Another fallacy.

Some economists realize the latter and try to adjust for it. One example is Robert Shiller, a Yale University Economics professor who realized earnings tend to fluctuate wildly near economic peaks and troughs—they tend to reach cyclical highs before market peaks, then tank during recessions. In Shiller’s view, this means the P/E ratio at a given moment might not reflect the market’s inherent valuation—it might be skewed by the economic cycle. On the one hand, a high P/E might indicate a later-stage bull market, when stock prices typically rise more than earnings. But it could also indicate a market trough: Profits are depressed after a recession, so even though stocks plummet during a bear market, P/Es can be high. Those who assume then that high P/Es mean an overvalued market get left behind when stocks recover.

In an attempt to account for this, Shiller developed a variation—the cyclically adjusted price-to-earnings (CAPE) ratio. Instead of using the most recent or future expected 12 months of earnings, CAPE instead uses the average of the last 10 years of corporate profits, adjusted for inflation. By smoothing out earnings variations, Shiller considers CAPE to be a more accurate indicator of the market’s true valuation—prices relative to long-term average profitability. Currently, CAPE (calculated as Shiller does) is signaling an overvalued market and, according to its proponents, below average returns over the next 10 years. Is this an omen markets are due for an extended period of poor returns?

There are times when a high CAPE has preceded poor future 10-year returns—1929 and 1999 being two. But there have been many other periods when CAPE was moderately above its long-term average (as it is currently) and stocks appreciated nicely over the following 10 years. The early 1960s and mid 1990s are two examples. Shiller himself even concludes the CAPE accurately forecasted only 40% of future 10-year stock returns. So even if you presume Shiller’s methodology has power, you accept the odds are worse than a coin flip it will accurately predict future 10-year returns at any given time.

But the methodology behind CAPE, in my view, is fundamentally flawed. Attempting to remove the influence of the economic cycle on profitability is a very curious practice, indeed. Consider the current CAPE. The last 10 years included one of the worst recessions in modern history, which skews the last 10 years of earnings lower. But why would earnings during the recession at all influence firms’ profitability looking forward? The conditions are different. Firms’ forward-looking profitability depends on the conditions and variables arising over the period ahead. Past earnings and price movement tell you nothing about that! Corporate profits could very well be much better moving forward—especially if the global economy reaccelerates as we expect. CAPE, like P/Es in general, is backward looking and tells you very little about the future.

Even if CAPE were an accurate indicator of stocks’ 10-year returns, it wouldn’t tell you how the market will arrive at these returns. Stocks could appreciate strongly over the next seven years, then experience a prolonged bear market, or vice versa. Periods with lackluster 10-year returns, like the last decade, can have perfectly fine runs in the interim—like stocks did from 2003 through 2007. Most investors would likely want their assets to participate in the good years, capturing those opportunities, and try to avoid at least some of the bad years. CAPE would proffer no clues for how to do this.

This isn’t to say P/Es can’t help forecast market trends—they’re very useful indicators of investor sentiment. P/E fluctuations during a bull market tell you how much of a premium investors place on stocks’ future earnings. Today, 12-month forward P/Es are modest, below their long-term average, even though stocks have had a huge run-up in prices since the bull market began. This suggests investors still aren’t willing to pay a ton for a share in stocks’ future earnings, even though fundamentals broadly are positive—investors are still quite skeptical. This means euphoric sentiment, which tends to characterize the final stage of the bull, is quite a ways off, implying the bull has further to run.

There are other indicators which, collectively, better forecast market trends, though admittedly, these are not helpful in forecasting 10-year returns—an impossibility, in our view. The difference between short- and long- term interest rates—the yield spread—is a powerful indicator of future economic growth. Political drivers, too, influence stocks—one key driver is the likelihood of major legislation being passed. No one indicator is the holy grail for forecasting broad market trends as a multitude of factors drive future market returns—but when the preponderance of the evidence points in a certain direction, it will seldom lead you astray.

CAPE, however, might. Using it to dictate portfolio decisions doesn’t seem like an effective means for outperforming markets over time. Investors would be better served to attempt to forecast markets over shorter periods (12 to 18 months), with economic, political and sentiment fundamentals their guide. CAPE might say this is a good time to get out of stocks, but investors who listen could easily miss a strong bull market. With fundamentals currently very positive, this bull market should continue for the foreseeable future.