Risk Increases Steadily As Stock Valuations Rise

Updated on

P/E 10 level AKA the Shiller PE and the S&P 500 Valuation-Informed Indexing #354

By Rob Bennett

Get The Timeless Reading eBook in PDF

Get the entire 10-part series on Timeless Reading in PDF. Save it to your desktop, read it on your tablet, or email to your colleagues.

The common way to think about stock valuations is to believe that they create bubbles that someday pop.  Investors who think this way are inclined to ignore valuations until they reach the bubble stage and then to try to guess the point at which they will pop with the aim of avoiding the worst effects of the overvaluation that caused the popping bubble.

I think it is a terrible mistake to frame the issue this way.

My view is that all increases in valuations increase risk by roughly the same amount but it is never possible to know in advance the P/E 10 level value that will trigger a price crash. Crashes are an irrational phenomenon (as is the overvaluation that causes them). So it is not possible to know in advance when a crash will come or to reliably avoid the effects of crashes through short-term timing decisions. However, it is possible to say when risk is elevated; higher P/E 10 level levels signal greater risk. So long as the investor aims to keep his risk profile constant by lowering his stock allocation in response to big increases in risk, he will be able to avoid most of the negative effects of crashes without ever even trying to jump out of the way of them.

Where did this bubble idea even come from anyway? Why is it that we are so drawn to a model of understanding valuations in which most valuation increases are harmless and only a few rare “bubble” cases that need to be dodged?

The answer is that bubble pops impress themselves upon the investor’s consciousness. One day he has a portfolio of x value and the next day he has a portfolio value of one-half x value. Investors notice that sort of thing. Even investors who are convinced that it is not possible to time the market worry about the pop of bubbles. So when most of us hear the word “ valuations”, our minds jump to the bubble issue and what to do about it.

P/E 10 level - is Eugene Fama right?

I worry about valuations a great deal. I don’t worry about bubbles at all. In fact, I would be inclined to agree with Eugene Fama, who has said that he does not believe that bubbles exist. I think that’s right. I believe that the bubble phenomenon is mythical. Overvaluation exists. Overvaluation is a big deal. But there is no knowable point at which harmless overvaluation becomes transformed into a dangerous bubble. All overvaluation is equally harmful.

The historical return data supports this way of looking at things. Say that the P/E 10 level has increased from 8 to 15. Stocks are priced at fair value when the P/E 10 level is 15. So no reasonable argument can be made that stocks prices have entered bubble territory when the P./E10 level reaches 15. But a regression analysis of the historical return data shows that the likely 10-year annualized return on stocks drops dramatically as the P/E 10 level level moves from 8 to 15. The most likely annualized return for a stock purchase made when the P/E 10 level is 8 is 15 percent real. When the P/E 10 level is 15, the most likely return is 6.5 percent real. That’s a big change. Valuations matter a lot when stock prices are nowhere close to bubble territory.

And I don’t see much evidence that they matter more when price levels commonly thought of as bubble levels are attained. If I were forced to identify a P/E 10 level that I view as the beginning point of bubble territory, I would pick the P/E 10 level of 25. We have never reached 25 and not seen a price crash follow within a few years. I believe that this is why Robert Shiller said in his famous Federal Reserve testimony of late 1996 (the testimony that led to Alan Greenspan’s coining of the term “irrational exuberance”) that investors who were heavily invested in stocks would live to regret it within 10 years.

The reality is that the crash anticipated by Shiller did not arrive until September 2008, 12 years following his prediction. And the three years immediately following the prediction produced some of the most amazing returns we have ever seen for the stock asset class. If we use a P/E 10 level of 25 as our indicator that we are entering bubble territory, many investors are going to conclude that bubbles are not such a bad thing.

Shiller’s warning becomes meaningful, however, if we stop thinking in terms of bubbles and focus instead on how each upward tick in valuations increases stock investing risk. Shiller was right that stocks became a highly risky investing class in 1996 and the fantastic returns obtained in the remaining years of the decade didn't change that; it is often the case that high-risk behavior does not produce an immediate negative outcome. We of course saw that outcome in 2008. And those who took comfort in the price jump that quickly followed miss the point if the risk-focused model for understanding the effects of high valuations is accurate. The P/E 10 level is now back at levels above 25, suggesting that stocks are again as dangerous an investment choice after the pain we experienced during the crash as they were before it came along.

Stocks are virtually risk free when the P/E 10 level is 8. The asset class has never produced poor long-term results beginning from that price point. But we have seen a P/E10 of 5. It is possible to buy at 8 and see a few years of poor returns. Not because a bubble is popping, but just because short-term returns are always unpredictable.

Risk is sky high when the P/E 10 level is 25 or higher. But it is entirely possible to achieve several years of strong returns on stocks purchase at that price level. Applying the word “bubble” to some price points doesn’t convey much meaningful information because, again, short-term returns are always unpredictble.

Risk increases with each uptick in stock prices. There are no special levels at which valuation increases cause greater increases in risk than all other price levels. Higher valuation levels are of course worse. But there is no reason to believe that the incremental increase in risk is greater at any particular set of price points.

Rob’s bio is here.

Leave a Comment