Yesterday, we talked about why future economic growth may well be at lower levels than it has in the past. Today, I want to look at what that might mean for future stock market returns.
It’s important to note that we’re not talking about the absolute level of growth but the difference from historical levels. This matters because, if the future is different from the past, then metrics that are based on past performance—such as stock market valuation levels—may also be different.
There are two questions to consider. First, is the market reasonably valued currently? Second, what will future returns look like from here? The first question is more subtle than it seems, as it depends on which value indicator you select and whether that indicator is appropriate under current conditions. That’s what we’ll focus on today.
The most common market valuation indicator is price to earnings, or how much the market will pay for a stream of future cash flows. I have written before about where those earnings come from, but now we’ll talk about what the P/E ratio actually implies. For this part of the discussion, I will use the standard P/E ratio, based on earnings for the past 12 months.
If the current P/E ratio is at historical levels, and is fairly valued, it implicitly assumes that conditions over the past 12 months are normal, historically. It assumes, for example, that unprecedented global financial stimulus by central banks, a U.S. housing recovery from the depths of an industry depression, historically rock-bottom interest rates, and substantial productivity gains as companies continue to keep their workforces at low levels—all of which led to record-high profit margins—are normal.
If we assume that the market is fairly valued, we’re essentially saying that all of these factors will continue to improve into the indefinite future. We expect central banks to not only continue to keep rates low but to continue—and even expand—quantitative easing. We expect companies to keep hiring at minimal levels, a trend that’s already beginning to turn. We expect to see housing values continue to rack up indefinite gains, and corporate profit margins to continue to improve.
If, in fact, these assumptions prove inaccurate, we will be paying too much today. The implicit assumption in the “fairly valued” scenario is that the future will look like the past—and, in this case, the past is the past 12 months, which I have described above.
One way to get past the idea of the past 12 months as the template for the future is to extend the look-back period. As you stretch out the time period, the influence of any individual factor, such as the Fed stimulus, declines. One such indicator is the trailing 10-year price/earnings ratio, also known as the cyclically adjusted P/E ratio, or Shiller P/E ratio, which uses 10 years of earnings data rather than one.
The argument for the one-year P/E is that it represents the most recent data, while the argument for the Shiller P/E is that it incorporates an entire business cycle. I prefer the Shiller P/E, as I don’t believe the past 12 months are the best indicator of the next 10 years.
If you look at them today, though, they give very different results as to whether the market is fairly valued. As of the start of May, for example, the trailing 12-month P/E ratio was 17.5—about 15 percent above the average of 15 since the 1870s, but reasonably consistent with the average over the past couple of decades. Overall, this figure suggests the market is not significantly overvalued. The P/E10, on the other hand, at 22.5 as of the start of May, is 36 percent above the average of 16.5—a significant difference.
The P/E represents one set of assumptions, which I have outlined above. If, however, you want to use the P/E10, you assume that we will have at least one systemic financial crisis, two deep recessions, and a housing market collapse. I certainly hope that’s not correct either.
The truth is probably somewhere in between, which means the market is moderately overvalued—which, in turn, means future returns will be lower than they would be if the market were cheaper. That’s not only reasonable, it’s common sense.
“Moderately overvalued,” though, actually begs the question—moderately overvalued compared with what? This is where lower future growth estimates come in. I will extend this conclusion to the next question, what this implies for future market returns, in tomorrow’s post.