Tugging on Valuations’ CAPE

Despite Tuesday’s rockiness and fearful headlines, stocks’ nearly six-and-a-half month rally from October’s low looks increasingly like a new bull market. A big reason? Alongside stocks’ strong returns, sentiment remains down in the dumps. Few believe in the rally’s staying power, with many hunting for reasons to justify their bearish stance. One such reason making the rounds Tuesday was the cyclically adjusted P/E ratio—aka CAPE—which is allegedly too high, signaling stocks are overvalued. We don’t buy it. Not only are valuations backward-looking, but CAPE is a fundamentally flawed indicator that doesn’t even try to predict near-term returns.

One big criticism of traditional P/Es, which compare stock prices to either the trailing 12 months’ earnings or analysts’ expectations for the next year’s, is that they are subject to too much skew from recessions and bear markets—a fair point. During bear markets, stocks typically fall before earnings do—and begin recovering while earnings are still falling. This tends to inflate P/Es early in a bull market, as you get a higher price divided by a much lower denominator—a curious math quirk that delivers a result many find counterintuitive (high P/Es around historically great times to own stocks). But it just means stocks pre-priced future corporate profits. As earnings bounce, P/Es usually even out.

CAPE tries to solve for this by comparing stock prices to the last 10 years’ worth of inflation-adjusted earnings, creating an alleged super valuation. The logic: By averaging out 10 years’ worth of profits, you compare prices to a more stable, allegedly accurate measure of core profitability. But in our view, that makes the measure super-duper backward-looking. It meant that until 2019, CAPE included depressed earnings from the financial crisis, which weren’t exactly relevant to stock prices on the eve of COVID. It also meant that in the 2000s, CAPE included the sky-high earnings at the end of the 1990s boom, which were similarly irrelevant at the time.

Adjusting for inflation adds more trouble. When it comes to economic data, we think inflation-adjusting is a crucial step to putting data in context and comparing across history as well as countries. But another cardinal rule to data analysis is keeping like with like. CAPE doesn’t. Its earnings are inflation-adjusted, but it uses nominal stock prices. That means its denominator is always adjusted lower, skewing CAPE higher. Inquiring minds might ask whether inflation-adjusting stock prices is a viable solution for this, but we don’t think so. Inflation-adjusted or “real” returns are an imaginary, largely academic construct. In the real (ha)^[i] world, investors earn nominal returns in hopes of outstripping inflation, and they pay taxes on those nominal returns. It isn’t even clear what the most logical deflator would be. CAPE uses the consumer price index to deflate earnings, but that seems suspect considering the sheer amount of companies that sell to other businesses, rather than consumers. The producer price index might be more relevant to their revenues, as it is to the vast majority of companies’ costs. Or you could get more granular and use the subsets that are most relevant to the individual companies. Or you could use the GDP deflator. Or or or.

Lastly, looking at CAPE’s history, there is clearly no magic level that signals a bull or bear market. (Exhibit 1) Nor is there designed to be one. CAPE doesn’t claim to predict returns over the next year. Nor does it claim to identify whether a bull or bear market is immediately ahead. Its goal—which we think is impossible—is to predict average returns over the next decade. Even if it were right, which it often isn’t, knowing average returns over a decade won’t tell you when and how those returns occur. A weak decade could include a very nice bull market or two, sandwiched between bear markets. CAPE won’t tell you when those cycles will turn, so it won’t help you navigate them. Note, too, that today’s allegedly high level matches levels seen smack in the middle of the 1990s’ and 2010s’ long bull markets.

Exhibit 1: A Long Look at CAPE

Source: Yale University, as of 4/25/2023.

Just looking at pure market predicting power, we don’t think CAPE is useful for investors trying to navigate stocks’ ups and downs. It has been high in bull and bear markets alike … and low in bull and bear markets alike. Its extra backward-looking nature doesn’t even make it a great sentiment indicator, unlike traditional P/Es. It is just a bizarrely constructed, awkwardly inflation-adjusted, occasionally entertaining, backward-looking thing.