To news & knowledge overview

Everything of value is defenceless

 Salomons Judgement

Whether the price that investors pay for American shares is too high is a widely discussed topic. It really depends on the way you look at it. Dusting off my own research nevertheless made me worry about the long term.

It’s ten years after the Great Recession. In my previous column, I wrote that a new recession seems rather more distant than closer, due to the rising interest rates and a steeper curve. That is why this column is not about the economy, but about another favourite subject: valuations. Specially for this occasion, I dusted off my thesis on predicting equity returns, and picked up the famous data set from Nobel laureate Robert Shiller, which goes back to 1871.


Let us first briefly refresh our memory. The P/E ratio of shares in the long term always returns to the historical average. It’s a slow process that statisticians call mean reversion. Mean reversion in P/E ratios implies that either the price (P, return) or the profit (E, growth) drives the ratio towards trend. A combination of both is – obviously – also possible. Investors are sometimes prepared to pay higher valuations for shares, but history shows that the subsequent years are accompanied by disappointing returns. The expected profit growth is always an illusion.


The valuations of American shares have again risen sharply, even if this is still less than anticipated considering the expected profits. Data from Standard & Poors indicates that in the third quarter of 2019 the expected operational profit is equal to $171, a P/E of 17x. Shiller has taught us, however, that it’s better to look at the reported earning. That gives us earnings of $131, or a P/E of 22x. But preferably we should also evaluate stated earnings over a longer (ten year) period. This ensures a realistic market valuation by restraining the effects of the economic cycle. Hence the name Cyclically Adjusted Price Earnings (CAPE) ratio. 


Expensive stock markets

American stock markets are again not cheap. Although the standard P/E is rather high, the CAPE is 31.5x – far above the average of 16.5x. It has only been so extreme twice before: in the roaring twenties and during the internet bubble. If history is any guarantee for the future, this will lead to an expectation of no more than 2 to 3% return for the next decade. The good news is that the CAPE only helps to forecast the long-term returns and doesn’t say anything about the next year. Mean reversion in P/E-ratios is, as said before, a slow process and the American economy is still running at full speed at the moment.  


In past years, Shiller had to endure a lot of criticism. His forecast was said to be too conservative. For the inventor, that was a reason to come up with an alternative version. He made no adjustments for secularly low interest rates or high profit margins. He only corrected for the fact that companies are increasingly buying their own shares instead of paying out dividends. And I think that’s fair.


However, the conclusion remains unchanged. American shares are also expensive based on the adjusted CAPE. It is not surprising that more and more investors in American bonds see a reasonable alternative for the coming years. A reassurance is that everything looks different in the short term and that valuations are much more attractive in many other regions than in the United States.

The author

Roelof Salomons

Subscribe to our letter

Leave your email address and be the first to receive the latest editions