How Not to Navigate Earnings Season

Mark Hulbert says what you need to know about earnings season has nothing to do with earnings. It’s about how the P/E ratio gets calculated.
Publish date:

The most important thing you need to know to make sense of earnings season has nothing to do with the earnings themselves.

More important than earnings, believe it or not, is how the price-to-earnings ratio gets calculated. There are many different ways of calculating that ratio, and though there is no one right away, it’s important to be consistent.

Unfortunately, few of us are. We instead calculate the P/E in one way and then compare it to historical values calculated in a different way. This sleight of hand can have a bigger effect on P/E ratios than even the most disappointing or encouraging of earnings reports.

Here’s why: Assuming you’re like almost everyone else on Wall Street, your primary takeaway from the upcoming earnings season will be the outlook for the coming year’s earnings. And you will calculate the market’s P/E ratio using that earnings estimate -- a so-called forward-looking P/E.

So far so good.

But if you’re also like almost everyone else on Wall Street, you will proceed to compare this forward-looking ratio to an average of historical P/Es. And that’s where the sleight of hand takes place: Historical P/Es are almost always calculated using trailing 12-month earnings. And since forward-looking P/Es are reliably lower than trailing ones, such a comparison makes the market look less overvalued than it really is.

It’s not an apples-to-apples comparison, in other words.

Consider the S&P 500’s forward-looking P/E according to earnings estimates from Birinyi & Associates: This ratio currently is 19.7. That compares to an historical average P/E over the past 50 years of 19.4, according to data from Yale University’s Robert Shiller -- suggesting that the market currently is, for all intents and purposes, fairly valued.

Because these historical P/Es were calculated using trailing 12-month earnings, however, the appropriate comparison is with the S&P 500’s current trailing 12-month P/E. That trailing P/E is 25.5, according to Birinyi & Associates, 31% higher than the 19.4 average in the Shiller database. That suggests significant overvaluation.

Research conducted more than a decade ago by Cliff Asness, founder of AQR Capital Management, and Anne Casscells of Aetos Capital, was the first I know of to point out this sleight of hand of which most of us are guilty.

They painstaking calculated what past P/Es would have been using forward-looking estimates rather than trailing 12-month earnings; their estimate is that the median of past P/Es would have been about 25% lower than when focusing on forward-looking rather than trailing P/Es.

No wonder the bulls would rather compare today’s P/E ratio to historical averages calculated with trailing P/Es than with forward-looking P/Es.

So be my guest and focus on forward-looking P/E ratios during the upcoming earnings season. There’s nothing inherently wrong with doing that.

Just don’t compare such ratios to historical P/Es calculated with trailing earnings.