
How can you tell if stocks are expensive?
Of course there is a thousand different ratios, indicators and indexes that attempt to measure the value in today’s stock market. Today we are going to look at one of the most popular ratios in particular, Shiller’s PE10 ratio, also known as CAPE ratio. What information can we get from looking at 100+ years of history between stock returns, treasury yields and CAPE ratio readings?
What is a P/E or CAPE Ratio?
A quick primer on P/E and CAPE for those unaware…For those that are familiar with a P/E Ratio, skip down to the next bold headline “CAPE Ratio-What is it?”
CAPE ratio stands for Cyclically Adjusted Price to Earnings ratio. You are likely familiar to the simple P/E ratio, which is calculated by taking a stock’s price and dividing it by its earnings. For example, take a look at Intel’s page on Google Finance:

Price = 42.98. Earnings Per Share = 2.85. So the P/E ratio is: 42.98/2.85 = 15.08. Just about what Google has listed (the small difference is due to intraday changes in Intel’s stock price. Google’s calculation lags a bit as prices change)
You can calculate the P/E ratio of the stock market as a whole as well. The S&P 500 is currently at 2,638. You can add up all of the earnings of the companies that make up the S&P 500 and get an earnings per “share” of the index just like we did for Intel above. Thankfully, people have done this calculation for you. Robert Shiller, Nobel prize winning economist, has a great site with a ton of up to date data including the earnings of the S&P 500 index: http://www.multpl.com/s-p-500-earnings/
As I write this, “Earnings per Share” for the S&P 500 is 104.75
So we can take 2,638 and divide by 104.75, and we get 25.18. That is the P/E ratio of the S&P 500. What does this mean?
When you buy the S&P 500 today, you are buying it at a 25 multiple of its profits. If you own a gas station, and that gas station makes a net profit of $100,000 per year, buying it at a 25 multiple means buying it at ~$2.5 million. If profits stay the same, it will take 25 years to “get your money back” from your purchase.
Is that a good trade off? That’s up for you to decide. But obviously the lower the P/E ratio is, the better, since we “get our money back” faster.
And a P/E ratio of 25 is starting to get expensive for the market as a whole. Here is a chart of the S&P 500’s P/E ratio since the late 1800’s:

A high multiple means that stocks are expensive compared to historical norms. Does that mean stock returns will be bad in the future? We’ll look into that later. First, what is the CAPE ratio, and how is it different from a simple P/E ratio?
CAPE Ratio – What is it?
Robert Shiller came up with an idea to help smooth out the chart you see above. He created the “cyclically adjusted P/E ratio”, which averages the last 10 years of earnings. So instead of dividing the S&P 500 price by 104.75 like we did above, we would instead divide by the average earnings over the last 10 years.
Once again, Shiller provides all the data you need on his site, and here is what the market’s “CAPE” ratio looks like over time:

Which makes the market look even more expensive!
This is a very widely followed indicator, and lots of people have been worried that it is a sign that future stock market returns will be low.
But has CAPE actually been successful in predicting future stock market moves? Here is CAPE vs subsequent annualized 10 year stock market returns:

It has actually done a very good job over a long period of time! You can see that when CAPE (the blue line) is high, future returns are low, and vice versa.
But the argument for high stock prices today is because interest rates are so low. How do we factor that into our analysis?
Another Way to Look at the P/E and CAPE Ratios
When you have a company’s P/E ratio, you have a “multiple” that you are paying for the company’s earnings. Likewise you can also use the P/E to calculate the stock’s “earnings yield” by inverting the P/E ratio to E/P, or Earnings to Price.
So with Intel above, we had a P/E ratio of 15.08. if we take 1/15.08, we get 0.0663, or 6.63%. Investors can think of this as a “yield” that Intel generates for its owners.
We can do this with the general market as well. The S&P 500 had a P/E ratio of 25.38, giving us an earnings yield of about 3.94%.
We can do the same with Shiller’s CAPE. Currently, the CAPE ratio is at 32.01. If we take 1/32.01, we get a “CAPE earnings yield” of 3.12%.
Inverting the P/E ratio doesn’t tell us anything new by itself. A low P/E ratio will mean a higher earning yield number, telling investors that there is more value in stocks at that price.
But it does give us a better way to compare the price and “yield” of stocks with other instruments, such as a bond, which is usually purchased for it’s yield.
A chart of historic CAPE Yield looks like this:

And here is the future 10 year returns on top of CAPE yield:

And the CAPE yield adjusted for inflation. To calculate, if the CAPE yield is 3%, and inflation was 1%, the Real Cape Yield is 2%.

Just for fun, with total hindsight bias, the chart coincides with notable periods in stock market history. Green represents times that would have been great to buy, red marks periods right before large stock market corrections:

So is the Stock Market Expensive Today?
Many will be quick to answer yes to that question (and I don’t necessarily blame them). But in reality we have to look at the price of stocks vs. other assets. Many investors are saying that with yields so low, stocks should be trading at a higher multiple. If a 10 year treasury bond yields 2%, then if stocks yield just 3%, even if that is low, it is better.
Now that we have calculated the S&P 500’s “Earnings Yield” and “CAPE Earnings Yield”, how does that compare to a treasury bond today?
It seems fitting that if we are looking at a company’s P/E ratio with earnings averaged over 10 years, then a 10 year treasury bond is a good comparison.
Currently, a 10 year treasury bond yields 2.33%
So the S&P 500 “yields” 3.12% based on the CAPE Ratio, and a 10-year treasury bond yields 2.33%.
Now, most will tell you that stocks should trade at some sort of premium to a risk-free treasury bond. And I think they’re right. But how much of a premium?
For the rest of this article I am going to use the term “Yield Premium” to describe the difference in the S&P 500 CAPE Earnings Yield vs a 10 year treasury bond yield.
Currently, with the S&P 500 CAPE earnings yield at 3.12%, and the treasury bond at 2.33%, stocks have a 0.79% “Yield Premium” to bonds.
I wanted to look how that compared to historical levels, here’s what I found:

And here is future 10 year stock market returns (annualized rate) on top of the CAPE yield premium:

Does this tell us anything?
If stocks have a high “earnings yield” (green line above), they are cheap and should have good long term returns (orange line). If stocks have a much greater earnings yield than the bond yields, they provide much more value right? For much of history, the CAPE yield premium was a great indicator of future stock market returns. Look at how peaks and troughs in the chart coincide with important times in stock market history:
- The peak in ~1919 marks the start of an incredible 10 year return in stocks. The S&P 500 (which didn’t technically exist at the time, but Shiller has prices going back as far as the late 1800’s by “recreating the index”) more than tripled from 1919 to 1929, with an annualized return of more than 12% for the decade.
- The chart also nails the 1929 peak. At a time when stocks were incredibly expensive (earnings yield was equal to the bond yield), it made no sense for investors to be in stocks, because bonds were offering an equal yield with no risk.
- The 1973 bear market is an obvious bottom in the chart, with a strong peak just a few years later after the market had a 50% correction.
- The tech bubble is nailed as well.
- Bottom of the great recession is notable peak.
So I saw these first couple obvious peaks and bottoms and thought I was on to something. But a big part of the chart doesn’t make sense…
The entire 80’s-2000 bull market occurred when stocks were “expensive” relative to bonds. Obviously, getting out of stocks in 1981, 1989, 1995, etc. was not a great idea. At a time when the CAPE yield was its most negative reading IN HISTORY, stocks returned 15% per year for the next decade! So much for a reliable indicator!
The chart above adjusted for inflation:
Is CAPE a Useful Indicator?
CAPE was a great indicator for most of history.
CAPE yield premium was a great indicator for most of history, but would have completely failed investors during the 80’s and 90’s.
I think one of the first charts in this article was the best, showing specifically that high CAPE ratios lead to lower future returns:

And although the 10-year returns stop in 2007, since we don’t yet have 10 year returns for 2008 and on, we can estimate future returns since some years have already happened. So what if the stock market returns 6% in 2018 and the years after?
A large part of the “bump” in returns due to the 2008 great recession has already happened. Unless 2018 has incredibly high or low returns, the beginning of the dashed green line is all but certain.
And here is the same future 6% returns over the CAPE Yield:

Here is the same future 6% annual returns over the CAPE Yield premium chart:

Like any indicator, CAPE is far from perfect.
If the CAPE ratio does have any prediction power, future stock returns will be low. A CAPE yield below 5% has never produced positive annualized 10 year future stock returns, except for 1995-1996.
Take that with a grain of salt if you wish, but I think that it should cause some worry for investors today. What do you think?