Don’t Be A Victim of Recency Bias
Is it possible that stocks aren’t overpriced? Financial adviser Josh Brown raises the possibility, arguing that earnings can grow into their prices. After all, Amazon, Netflix, and Nvidia have seemed overpriced to investors for a long time, but their economic performance keeps improving. As Brown puts it, with all of these stocks in the recent past, “[t]he fundamental stories grew up to justify the valuations investors had already been paying (Brown’s emphasis).”
And this can also happen to entire markets. Five years ago, the market’s cyclically adjusted P/E ratio (CAPE or Shiller PE) was higher than it had been in 87% of all readings up until that point. But the stock market has been up 90% since then. “No one could have known that the fundamentals would arrive to back up the elevated valuations for stocks eventually,” according to Brown.
This last statement is odd. In May of 2013, the S&P 500 carried a CAPE of 23. Now its CAPE is 31. It’s not clear from this simple valuation metric that stock earnings have grown into their new, elevated prices. Past ten-years’ worth of earnings ending in 2013 were $78, according to Robert Shiller’s data. For the most recent ten-year period, they are $84. Ironically, one could make the argument that earnings have grown into the 2013 price five years later, but not the 2018 price. It we apply the May 2013 price to the past decade’s worth of earnings ending today, we get a CAPE of around 21. That’s much more reasonable than the current one of 31.
In fact, if we agree that the long-term historical average CAPE of nearly 17 is outdated, and that the new average should be around 20 or 22, then the 2013 price of the market relative to the past decade’s worth of earnings ending today is roughly the correct valuation. That also means all the price advances the market has made since 2013 do not reflect underlying economic reality or earnings power value of the market. In other words, earnings have increased, but stock prices have increased much more so that the market should be trading at 2013 prices given the past decade’s worth of earnings.
Brown’s point, of course, is that the earnings growth of the past decade can repeat over the following decade. But that also means that for stocks to deliver robust returns, the current 31 CAPE valuation must reappear 10 years from now. That’s possible, but investors and advisers must contemplate how they would like to bet and what they must tell clients if they are behaving as fiduciaries.
It’s possible that we could wake up to a 31 CAPE in a decade, and that U.S. stocks will have delivered 7% or so nominal returns (2% dividend yield plus 4%-5% EPS growth). It’s also possible that earnings-per-share can increase at a greater clip than they have historically. Nobody should say those things are simply impossible. But if you are managing your own money, or advising others in a fiduciary capacity (which means you must treat their money with all the care you do your own), how reasonable is it to expect that as what forecasters might call a “base case?” At best, assuming we’ll all wake up to a 31 CAPE in a decade must be a very rosy, low-probability scenario.
There’s an irony to Brown warning against those who carry on about backward looking valuation metrics. One of the most well-known observations of behavioral finance is that human beings can be seduced by recent patterns, including recent securities price movements. We tend to assume, without any evidence other than the recent pattern, that price trends will continue. Everyone will have to decide for themselves whether deriving encouragement from a 90% stock price move without a commensurate earnings increase, as Brown does, reflects proper attention to simple arithmetic or our susceptibility to extrapolate recent stock price movements and returns into the future.