Why Are We Still Using the P/E Ratio?

Art by Wesley Allsbrook

The price/earnings ratio assumed its prominence in the frothy market of the Jazz Age, when earnings supplanted dividend yields as the most important measure of a share’s worth.  Some wonder now whether P/E should go the way of the Charleston and other 1920s enthusiasms.

The price/earnings ratio for the trailing 12 months, or TTM, its most common iteration, remains in wide use. After all, earnings are the most important underpinning of corporate performance, and the P/E concept is easy to understand. “People need a simple number to make decisions,” said Doug Foreman, chief investment officer of Kayne Anderson Rudnick Investment Management.

Asset managers, i.e., the smart money, employ many other metrics in assessing market moves or individual stocks’ prospects. Price to book value, price to sales, and enterprise value to EBITDA are among the many other measurements that professionals enlist in their evaluations of stocks.

The P/E ratio benefits from historical precedent, having been at the forefront of stock valuations for the past 100 years. Throughout much of the previous three centuries, bonds were the primary investment product. After World War I and the full flowering of mass production and a rising middle class, equities became extremely important.

Aside from the dot-com interlude, when eyeballs per click were in vogue, the profits companies generate were the common-sense way to judge shares. As Benjamin Graham and David Dodd wrote in their 1934 classic book, Security Analysis, “common stocks have come to depend exclusively on the earnings exhibit.”

Still, the standard 12-month P/E has a few problems, As influential newsletter editor Lyn Alden put it: “The P/E ratio can be very misleading.” And how. The multiple can send false signals about the market, earnings can be slippery, and the ratio may wrongly brand some individual stocks as too cheap or too expensive. Of course, other yardsticks are far from perfect, too. Could it be that P/E is worthwhile, if you make a few allowances?

False Signals

The storied ratio is certainly deficient as an indicator that the market is too high and is poised for a fall. The stock market’s historical average P/E, according to Yale economics Professor (and Nobel Laureate) Robert Shiller, is 15.8, using data stretching back to the 1880s. For the S&P 500, the current average 12-month P/E is 22.2. Conventional wisdom is that, once the market gets too high, it will revert to the mean—in other words, to 15.8.

But that doesn’t happen too often, especially lately. Sure, the market, as embodied by the S&P 500, will tumble into the bear pit come the next recession, and stock prices and earnings will shrink. Yet absent an economic slump, which has been predicted for years as the current decade-old bull market rumbled upward, a market pratfall doesn’t appear imminent (absent a catastrophic surprise, of course, like a big war or a vast natural disaster).

Since 2000, the S&P 500’s P/E has been north of 20 most of the time. This century, on a yearly basis, it only touched 15 once, in 2011, amid the economy’s shaky emergence from the Great Recession. That raises an important question: Do data from the late 19th and early 20^th centuries make sense in today’s world. Probably not. Today’s higher valuations may be the new normal.

The stock market has changed enormously. “The market has become more liquid and more efficient,” said Norm Conley, CEO of JAG Capital. He emphasized that the new era started in 1975, with the removal of fixed commissions, which made trading cheaper and led to the rise of discount brokers like Charles Schwab. The advent of digital trading means trades happen faster and the market is available to all. In the 1990s, decimalization took effect: Stocks no longer were valued in one-eighth of a dollar, but in pennies.

The greater ease of trading means that, when the market is going up and the economy is doing well, share prices will vault ever higher amid the excitement.  Since World War II, despite some harrowing downdrafts, the market and the economy have ascended. Since stock prices anticipate the future, they often climb faster than recent earnings.

Since 2014, the yearly P/E average has been 20 or above. The S&P 500’s sharp correction in late 2018 (down 19.9%, a hair beneath the standard bear market level of a 20% drop) was reversed this year. And today’s P/E isn’t the highest. It is shy of this decade’s peak, 25 in 2017.

A more realistic average is 18.5, according to Sal Bruno, chief investment officer of IndexIQ, who builds portfolios by combining fundamental and quantitative research. Bruno got this number by calculating data from the last 25 years, on the sensible theory that the recent past will prove a better guide. Today’s S&P 500 P/E, therefore, “is elevated, though not so dangerous,” he said.

Slippery Earnings

Financial engineers have many techniques to make earnings look better than perhaps they should be. While these in most cases are legal, investors may end up bamboozled. But their P/E levels don’t jump too high or sink too low. The E is solid, advancing at a pace that somehow manages to exceed analysts’ expectations.

For example, a retailer might switch to first-in, first-out accounting, counting merchandise sold to consumers as the oldest—and less-costly for the company to buy from its wholesaler—items in inventory. That way, profit margins are higher than if the company used the last-in, first-out method, where the newest goods cost more for the company to acquire.

General Electric in the 1980s and 1990s had remarkably smooth earnings increases under its CEO at the time, Jack Welch. The company’s then-mighty financial arm, GE Capital, could be counted on to maneuver reserves around to fill in the dips. That pleased investors, who like nice, dependable results.

After Welch’s departure, GE’s profits dynamo began to falter, and some of its tactics became more aggressive. In 2009, GE agreed to pay a $50 million fine to the Securities and Exchange Commission by resorting to tricks like selling locomotives in a way that seemed like loans.

P/Es can be out of whack for reasons other than numbers manipulation. Consider 1999 to 2000, when faddish tech stocks saw their prices explode, even though these outfits had minimal earnings at best. By 2001, when reality had set in and prices declined, the S&P 500’s average looked oddly high, at 46. The same situation prevailed in 2008, the year of the financial crisis, with the P/E as 70. Both prices and earnings plummeted, albeit not at the same rate. “The E went down faster than the P,” recalled JAG’s Conley.

Attractive or Unattractive?

Low P/Es, or high ones, may indeed be misleading. Other attributes need to be weighed.

To some, low P/Es are a virtue. Value investors are always on the hunt for stocks with unjustifiably low P/Es. These can disappoint, naturally. Clothing merchant Gap (P/E: 7), for instance, is a darling of many value players who believe its sales slump is over-done. Well, its stock price has fallen by more than half over the past five years amid erratic earnings and sluggish sales.

On the other hand, as Kayne Anderson Rudnick’s Foreman noted, energy stocks were in the dumps in the early part of the century, and then rose nicely. While they are off lately, just compare Exxon’s October 2003 price ($34) to today’s $69. In 2014, the oil giant changed hands for $101.

Even more puzzling is the criticism of some leading high P/E stocks as overvalued. Are the tech stars that have propelled much of the market’s advance in recent times, the FAANG stocks Facebook (P/E: 31), Apple (20), Amazon (71), Netflix (112), and Google parent Alphabet (24), indeed too pricey? Legitimate worry exists about Netflix’s staying power, but the rest are dominant in their fields.

Google was derided widely in its early existence as overpriced. In the middle of the last decade, its P/E was 60: The share price surged from $115 in September 2004 to $475 in January 2006. Today, it’s $1,220. When politicians are talking about breaking up a company for being too powerful, at least it shows the business is successful.

Other Alternatives

All that said, are there better ways to value stocks?

Dodd and Graham’s preferred metric was price to book value, not P/E. While that may have worked in the 1930s, book value (assets minus liabilities) is not what it once was. “The economy has structurally changed,” said Frank Rybinski, chief macro strategist at Aegon Asset Management. Especially with the advent of high-tech, a lot of intangibles have crept onto the balance sheet. The value of a brand name and the like may or may not be exactly precise numbers, to be charitable.

Price to sales is another oft-touted method, although that for the most part ignores earnings.  Earnings remain key to shareholders for good reason. What’s more, when the market cap is inflated, the comparison of price to sales can turn out to be no better than the P/E in gauging valuation.

Over the past three years, ride-hailing tyro Uber’s revenue has tripled. Anticipation was electric for the company’s initial public offering in May. Prior to the IPO, the price-to-sales ratio was a high 6 (1 to 2 is considered ideal). Alas, its red-ink earnings record caught up with Uber, and the elation fizzled—once again demonstrating the potency of earnings for investors. The stock launched at $45 per share and has lost a third of its value since.

Since the takeover era of the 1980s, enterprise value (equity plus debt) to earnings before interest, taxes, depreciation, and amortization has claimed a place at the valuation table. The virtue of this measure is that it encompasses debt, a vital corporate finance tool. Trouble is, ignoring costs like interest and taxes (even though clever accountants can delay or re-shape them) is missing key liabilities that affect corporate health.

Another alternative is to keep the P/E, and instead use estimated future earnings, rather than those from the previous 12 months. The downside here is that the future has an odd habit of making fools of mortals who attempt to divine it. Earnings forecasts tend to be on the optimistic side. The collective projection for S&P 500 companies for 2020 is to grow 10%. In a slowing economy, that seems overly sanguine. Last fall, analysts said earnings would grow 10% in 2019. With year-end in sight, that estimate has collapsed to 2%.

Yale Professor Shiller thinks that the best course is to revamp the P/E, which he argues is too shortsighted. Instead of going back 12 months, his Cyclically Adjusted PE Ratio, or CAPE, covers 10 years in the past, employing the average of inflation-adjusted earnings.  The CAPE is now at 28.9, far above the metric’s 16.9 long-term average. 

Criticisms of CAPE are that it is too pessimistic because, among other things, it still is using the end of the Great Recession when, the reasoning runs, prices and earnings were depressed to an anomalous extent. Stephen Wilcox, a finance professor at Minnesota State University, Mankato, argued in a 2011 paper, “the average business cycle has been significantly shorter than 10 years,” a condition that skews the CAPE.

The Bottom Line

The view of the market’s future now and the direction of valuations, as determined by P/E, is mixed. Interest rates, after three years of rising, are going down again.

That is propitious for stock prices: Uber-investor Warren Buffett once remarked that rates “act on financial valuations the way gravity acts on matter. The higher the rate, the greater the downward pull.” That’s because the return from risk-free Treasury paper becomes more appealing than that from stocks. When rates descend, the reverse is true. And if earnings are contracting and prices go up, then P/Es will move even higher.

So perhaps we should give two cheers to the traditional P/E, within reason.

Related Stories:

Shiller Doubts Booming Earnings, Now Propelling the Market, Will Last

The S&P 500 Keeps Singing the Sour Earnings Song

Is an Earnings Recession Really Coming?