Daily Ticker

Everything You Always Wanted to Know About P/E (But Were Afraid to Ask)

One of the most common numbers you hear if you listen to stock forecasts is the "P/E ratio"--the price of the stock or market relative to the company's earnings.

Sometimes, the P/E ratio can be very useful for getting a sense of whether a stock or market is cheap or expensive.

At other times, however, the number can be highly misleading.

One of the most common ways the P/E ratio can be misused, for example, is when investors cite the P/E of a small, rapidly growing company that currently has a low (or no) profit margin. In these cases, the company will often have an astronomically high P/E on its historical earnings, which can lead some investors to conclude that it is "ridiculously priced."

What these investors often don't realize, however, is that stocks trade based on future earnings expectations, not past earnings, and the company's earnings might be expected to grow extremely rapidly in the future. Investors who are looking at the stock's price relative to possible future earnings, therefore, might conclude that the stock is reasonably priced (or even cheap).

Another way the P/E ratio can be misleading is when it is applied to the stock market as a whole. Right now, for example, the S&P 500 has a P/E ratio of about 15-times this year's expected earnings. That's about an average P/E, so many people look at it and conclude that the market is reasonably priced.

The trouble is that the earnings of the S&P 500 are highly volatile, based on the profit margins of the underlying companies. When times are good, as they are now, profit margins are very high, producing very high earnings. At other times, however, profit margins collapse, producing very low earnings. This "mean reversion" of profit margins means that only considering a P/E based on a single year of earnings can trick you into thinking the market is cheap or expensive.

When profit margins are near peak levels, for example, as they appear to be now, the market's P/E will look artificially low. When profit margins are near trough levels, meanwhile, as they were in 2009, the market looks artificially expensive (recall the record-high P/E ratios at the depths of the financial crisis, when the market was actually cheap and was about to blast off on a four-year bull market).

The way to create a P/E ratio that is actually useful for predicting long-term stock market returns (not short-term ones) is to average earnings over many years, thus muting the impact of the business cycle. The professor who popularized this method, Professor Robert Shiller of Yale, calls this method the "cyclically adjusted price-earnings ratio," or CAPE, and it has been very good at predicting future 10-year returns over the past 120 years.

Right now, the market's CAPE is about 22X. That's not extraordinarily high--at the market's peak in 2000, the CAPE was 45X--but it's distinctly higher than average. Based on the past century, when the CAPE starts at this level, the next decade's return for the market has usually been low, about an average of 3% (inflation-adjusted) per year.

So, if you believe that past is prologue, you should have low expectations for long-term stock returns, regardless of how many people talk about the market's "average P/E."

But...

There is, of course, no guarantee that the CAPE ratio going forward will regress to the same "mean" as it has in the past. And some long-term stock-market bears, such as the folks at GMO in Boston, are now asking whether it's possible that corporate profit margins have permanently increased. If they have, stock returns over the next decade could be significantly higher than Professor Shiller's CAPE is suggesting.

We asked Professor Shiller about that.

He said that, while anything is possible, he doubts that anything has changed. Stock market forecasters, he notes, are forever arguing that "this time is different," and it almost never is.

So the safe bet is probably to conclude that stock-market returns over the next decade are likely to be considerably lower than the 7% per year average (inflation adjusted) of the last century.

View Comments (14)