Matt: What would you say are the strengths and weaknesses of looking at stock prices this way? And how is this different from what other people do?
Bill: I think that by keeping it to a 30-year period, you’re acknowledging that you aren’t including the rest of the data. And the rest of the history of the market unquestionably is one of lower valuations, in terms of multiples over most market measurement periods beginning in the 1870s up through 1988. There are a couple of market peaks — late 1920s, 1960s, in terms of multiples that are pricy as well— but mostly you’re looking at a data set outside of this range that makes the market look at the last 30 years as consistently expensive. That’s one limitation of the data.
On the other hand, a 30-year time period does tell you something. At least investors should consider whether there are things in the market that mean that a higher multiple, which has been more or less sustained for 25 to 30 years, is a function of things that are different in this market than was the case 70, 80, 90 years ago.
Matt: Before we jump to any conclusions about today’s prices, looking at the last 30 years, what jumps out at you? Is there anything about past market valuations that you learned that you didn’t already know?
Bill: I like to take the Silverblatt numbers, run them through an Excel heat map, and look at the colors. Instead of trying to say “this is what was happening,” “this was the excuse for this happening,” I just want to answer the question: “What does a simple math calculation look like?” There are times of bright, bright, bright red. That signifies elevated P/E numbers, and in theory that’s when you shouldn’t have been investing. And you can see that all over the late 1990s, starting about 1997 and really extending into 2002. And then there are periods of bright green, which signifies lower P/E numbers, and that includes late 2008 and all of 2009, of course, although many people thought the market was expensive then.
But there are a couple of periods where some indicators of the P/E ratio told you the market was expensive, and some indicators told you it was cheap, like the depths of 2008. If you were just looking at a simple GAAP price-to-earnings, the market looked like it had never been that expensive due to temporarily collapsed earnings – especially on GAAP earnings, less so on operating earnings. But if you’re looking at price-to-trend earnings, it looked extremely cheap. And if we were to end that experiment today, we would say, “Turns out trend earnings is what investors should have been looking at, rather than the GAAP earnings because if they did invest in late 2008, early 2009, they’ve done very well for themselves.”
Matt: What, of all these ways of looking at price to earnings, is the most predictive?
Bill: The price to the trend. And the way I’ve done this is take the beginning of the data set, the 1988 earnings, and run that through what is the growth, with that being the starting point? And the end of 2017 being the endpoint. You’ve got earnings per share compounding at a little bit above 6.2% annually, through the S&P 500. So there are times when earnings lagged a little bit, and there are times when they get ahead a little bit. And part of what I see in this data set is earnings getting ahead of that trend in the late 1990s. Earnings were growing faster than that, and investors got overly enthusiastic about extending that growth rate out ahead of themselves for a couple of decades. The earnings didn’t end up growing faster than 6% annually, even though they were temporarily growing a good deal faster than that.
That is a warning signal for today’s investors. We might be looking at the current earnings-per-share growth, this year versus last year, at somewhere around 28%, 29% growth. That’s absolutely not going to continue. Part of that has to do with tax cuts and the very real and permanent effects that the tax cuts have on earnings levels. And I calculate that as a one-time jump of 8.5% above the previously established trend. I think investors are going to get in trouble if they start extending their forecast of earnings growth out above about 6% from today’s levels though.