By Nir Kaissar
(Bloomberg Gadfly) --People are worried about U.S. stocks. The S&P 500 is up 251 percent since its post-financial crisis low on Mar. 9, 2009, and valuations look stretched.
One measure in particular is sending a warning signal. As Bloomberg contributor Ben Carlson recently pointed out, the cyclically adjusted price-to-earnings, or CAPE, ratio calculated by Yale professor Robert Shiller has reached 30 for only the third time since 1881 -- the first year for which the CAPE ratio is available. The other two occasions were just before the Great Crash in 1929 and in the run-up to the dot-com crash in 2000.
As Carlson correctly points out, comparisons with those two infamous market crashes, however warranted, are hardly proof another crash is imminent or even inevitable. But there’s a more dispiriting lesson lurking in the historical CAPE ratios. Namely, that returns from stocks tend to decrease over the medium term as the CAPE ratio rises.
The following chart shows the intersection of each monthly CAPE ratio since 1881 and the S&P 500’s subsequent 10-year annual return. (There’s no relationship between the CAPE ratio and subsequent short-term returns.) There are two inescapable conclusions from the data. First, the correlation between the CAPE ratio and subsequent returns is far from perfect. That’s why using the CAPE ratio or any other valuation metric to make binary, all-in-all-out market timing moves is likely to fail.
The second takeaway is that valuations have a meaningful impact on future returns. As valuations rise, the probability of high returns gradually diminishes and is eventually replaced with low or negative returns. A CAPE ratio of 30 or greater -- which is where we are now -- has historically been followed by low single-digit or negative 10-year annual returns.
The historical relationship between valuations and returns isn’t just a coincidence. Here’s how it works in real life: Say, for example, that a stock sells for $10 and has $1 of annual earnings. That investment has a P/E ratio of 10 and a yield of 10 percent ($1 divided by $10). A yield that high is likely to attract new investors, which raises the price of the stock -- and its P/E ratio -- and adds to the hefty 10 percent yield enjoyed by current shareholders.
But the valuation can only climb so high. Once the stock price reaches $30, let’s say, the P/E ratio jumps to 30 and the yield drops to just 3.3 percent. A yield that low isn’t likely to attract many investors, and some shareholders will inevitably sell the stock and chase a higher yield elsewhere. The resulting decline in the stock price leaves late-arriving shareholders with losses that further reduce their paltry 3.3 percent yield.
Despite this relationship between valuations and returns, there’s lots of debate about whether valuations should inform investing decisions. That’s the wrong question, however. The better question is whether valuations are a more useful tool than the information investors currently use to make investing decisions. I think the answer is yes.
For example, when ordinary investors shop for stock mutual funds and ETFs in their brokerage accounts or 401ks, they are customarily greeted with the funds’ 1-year, 3-year, 5-year, 10-year, and since-inception returns (and the funds’ expense ratios, if they’re lucky.) Having no other basis for choosing a fund, investors naturally pick the funds with the best recent performance.
But there’s no relationship between past returns and subsequent returns, as any fund prospectus will tell you. The following chart shows the intersection of the S&P 500’s trailing 10-year annual returns and subsequent 10-year annual returns since 1871. There’s no discernable relationship between the two.
Investors would be better informed if a fund’s earnings yield -- or some other valuation metric -- appeared alongside that fund’s recent performance. Consider the following table, which shows both recent performance and the normalized earnings yield (inverse of the CAPE ratio) for three popular indexes. I suspect investors would be much less likely to bet everything on U.S. stocks if they were alerted to the fact that U.S. stocks’ high recent returns are likely to compromise future returns.
Let’s acknowledge the link between valuations and investment outcomes. And then let’s give investors the valuation data they need to make fully informed decisions.
This column does not necessarily reflect the opinion of Bloomberg LP and its owners.
Nir Kaissar is a Bloomberg Gadfly columnist covering the markets. He is the founder of Unison Advisors, an asset management firm. He has worked as a lawyer at Sullivan & Cromwell and a consultant at Ernst & Young.