The stock market has been in a new price regime for 20 years

Sam Ro
Managing Editor

Almost every stock market forecaster will agree that stock prices look expensive when considered relative to earnings (P/E). Indeed, this is one of the main reasons why Wall Street strategists aren’t very bullish about the second half of 2017.

However, in two recent letters to investors, GMO’s Jeremy Grantham questions the way we often look at P/E ratios to think about the stock market. Specifically, he suggests the stock market may be — at least temporarily — in a “new normal” of elevated P/Es.

The P/E, or price/earnings, ratio is the most popular and most straightforward way of measuring value in the stock market. For the S&P 500 (^GSPC), it’s calculated by taking the index level (P) and dividing it by earnings per share (E). There are several ways to calculate this: You can use trailing 12-month earnings or the forecast for next-12-month earnings; you can use GAAP earnings or non-GAAP earnings; you can use one year’s worth of earnings or you can use the average of 10 years worth of earnings.

Analysts and investors draw the same conclusion from the P/E: If it is above some long-term average, then stocks are expensive. And vice versa.

As we begin the second half of 2017, it’s worth revisiting what Grantham has said.

What if the average P/E we’re all using is wrong?

Grantham considers the last 20 years of stock market behavior and observes that things aren’t what they used to be.

“For a long and painful 20 years – for someone betting on a steady, unchanging world order – the P/E ratio stayed high by 1935-1995 standards,” Grantham wrote in May. “It still oscillated the same as before, but was now around a much higher mean, 65% to 70% higher! This is not a trivial difference to investors, and 20 years is long enough to test the apocryphal but suitable Keynesian quote that the market can stay irrational longer than the investor can stay solvent.”

“Old normal” valuations versus “new normal” valuations.

The issue is not so much that average valuations seem elevated. Rather, it’s that traumatic market events haven’t jolted valuations back to the more historic levels.

“After the bursting of the tech bubble, the failure of the market in 2002 to go below trend even for a minute should have whispered that something was different,” Grantham said. “Although I noted the point at the time, I missed the full significance. Even in 2009, with the whole commercial world wobbling, the market went below trend for only six months. So, we have actually spent all of six months cumulatively below trend in the last 25 years! The behavior of the S&P 500 in 2002 might have been whispering in my ear, but surely this is now a shout? The market has been acting as if it is oscillating normally enough but around a much higher average P/E.”

Now, Grantham isn’t totally lost here. The veteran investor says elevated P/Es can be explained at least somewhat by low long-term interest rates, which have been in a bear market for nearly four decades. This is something his Warren Buffett emphasized at this year’s Berkshire Hathaway annual meeting.

Waiting for the old normal

Grantham doesn’t believe this “new normal” of higher P/Es will be permanent. But he does think it could drive markets for a long time, especially as interest rates only begin to rise from low levels.

“In short, I think lower rates than those in effect pre-1997 are likely to be with us for years, and the best we can hope for is several of the factors we’ve discussed moving slowly to push real rates higher,” he said. “In the meantime, while we wait for higher risk-free rates, investors – value managers included – should brace themselves for continued higher multiples than those of the old days.”

Grantham estimates this could be another 20 year journey. For now, investors shouldn’t be surprised if the next selloff again fails to bring the stock market back to historically cheap levels.

Jeremy Grantham

“If we have a bear market soon, even a severe one, will it recover to the new normal of 23x or the old normal of 15x? I believe the former,” Grantham said on Thursday. “Let me remind you that that new 20-year era of higher prices was very severely stress-tested by the 50% decline from the 2000 Tech Bubble and the 50% decline in 2008-2009, perhaps the biggest foul-up of the financial system in modern times. Neither spectacular event was enough, apparently, to jolt the price trend back to its former lower level.”

Be careful when you use P/Es

This discussion makes one think about what P/E really means. Literally, when you buy stocks at a 15 P/E, you’re paying a price that’s equal to 15 years worth of earnings. Is it really that much of a stretch to be paying for 23 years worth of earnings instead? Who really knows what the next 15 years or 23 years worth of earnings will actually look like?

And while we’re on the subject of averages, it’s important to recognize that today’s elevated P/Es are inputs for these calculations. If the P/E averages 23 for the next 80 years, then 80 years from now we’ll look back and say 23 was the 100-year average. That’s just how math works.

Let’s also not forget how incredibly difficult it is to forecast interest rates. As Buffett said, “Everything in valuation gets back to interest rates.” Low rates make the relative value of stocks to bonds more attractive, and they make the absolute value of stocks higher because discount rates are lower.

As with everything in stock market forecasting, the truth only becomes clear in hindsight. For now, Grantham’s proposition is yet another reminder that we should be wary of relying solely on P/E ratios to make short-term investment decisions.

Sam Ro is managing editor at Yahoo Finance.
Read more:

By using Yahoo you agree that Yahoo and partners may use Cookies for personalisation and other purposes