Bubble warning: The S&P 500 has only been this expensive for 4% of the past 140 years
On July 12, the S&P 500 posted a new all-time record close of 4385, extending its gains since the start of 2021 to 16.8%. But the market reached another milestone that should gravely concern prudent investors. A highly respected valuation gauge developed by Yale economist and Nobel laureate Robert Shiller hit a mark showing that the S&P 500 is now pricier than in 96% of all quarters over the past 141 years. Put differently, big-cap stocks have been this expensive only 4% of the time in the recorded history of equity markets.
The new reading is so extreme that it suggests stocks may be soaring in a speculative frenzy unhinged from fundamentals. It also spotlights the bulls’ view that the current regime of super-slim interest rates means that big-caps can sustain, or even expand, price/earnings (P/E) multiples virtually never before seen. The Shiller data points to a different outcome: Just because equities are likely to beat ultra-low-yielding Treasuries in the years ahead doesn’t mean stocks will keep delivering strong gains. Quite the contrary: The data suggests that Treasuries will do extremely poorly, and the S&P should underperform as well, only less so.
U.S. big-caps will battle three powerful headwinds. First, the fall in “real” rates that has done so much to lift valuations will inevitably reverse. Second, that downdraft pressures today’s high-flying P/E multiples. Third, the U.S. economy is destined to grow at a far slower pace than in recent decades, reducing profit gains from stupendous before the pandemic to slim in the years ahead. Since mid-2020, it’s been all about momentum. Going forward it will be all about the other Big M—the market Math.
The highest level in nearly a quarter-century
The yardstick flashing red is Shiller’s cyclically adjusted price/earnings ratio, or CAPE. Calculating the current P/E using the past year’s earnings per share (EPS) is misleading, because profits are extremely erratic. When earnings spike to unsustainable heights, P/Es shrink to make prices look artificially cheap. When EPS craters, as during the pandemic, shares appear outrageously expensive even though profits are certain to rebound, as indeed happened in the current recovery. Shiller’s CAPE removes those distortions by using a 10-year average of earnings per share, adjusted for inflation. (The numerator is simply the current “price” of the S&P 500 index.) By smoothing the peaks and valleys, the CAPE presents the best picture of what U.S. big-caps are capable of earning through the undulating cycles.
At the close of 4385 on July 12, the CAPE reached 38.39. To put that number in perspective, the reading just prior to the stock market crash of 1929 was one-sixth lower, at 32.56. In the 141 years for which Shiller has been measuring the CAPE, it exceeded 38.39 only in a single period: The “tech bubble” of late 1998 through 2000. The first time ever that the CAPE climbed past 38.39 was in December 1998, when it reached 38.82. It remained above that level, and went well into the 40s for the first and only time, for the following 23 months in the biggest postwar craze ever.
The CAPE fell below 40 in late 2000, and never again attained 38.39 until July 12.
Hence, as measured by the CAPE, the S&P has been more expensive only in 24 of the 542 quarters since 1881 recorded in Shiller’s database. That boils down to a ratio of four quarters out of 100, and until this week all of those quarters fell in single period when equities became notoriously overvalued.
At most times, a CAPE this size would signal big trouble.
In the past, what follows after the CAPE gets this lofty isn’t pretty. When the CAPE hit 38.82 in December 1998, the S&P stood at 1190. Five years later, in December 2003, it had shed 9.2% to 1081, and a decade hence in December 2008, the loss expanded to 27% at an S&P of 879. Put simply, a CAPE well over 38 is an off-the-charts outlier that, the single previous time it happened, led to horrible returns over the ensuing five and 10 years.
If you flip the Shiller PE, you get the CAPE “earnings yield.” That measures what companies can be counted on to earn for every dollar investors hold in their stock. At a CAPE of 38.39, that number is an extremely slender 2.6%. In most periods, the earnings yield is a good predictor of the returns investors can expect in the future. Add 2% inflation, and the CAPE is projecting something like gains of 4.6% going forward. That’s nothing like the double-digit bounty the S&P has been furnishing in recent years.
But in most periods, the CAPE earnings yield would also far overstate what investors can expect. That’s because the current CAPE is so elevated that it’s likely to fall sharply, erasing the gains from dividends and the impact of buybacks and rising profits.
A second Shiller measure, however, signals that today’s P/Es aren’t inflated at all. What appears to be huge valuations make sense because interest rates are so modest and likely to stay that way. Shiller’s alternative metric is his Excess CAPE Yield. It’s calculated by taking the regular CAPE earnings yield, and subtracting the “real” or inflation-adjusted rate on the 10-year Treasury. What the Excess CAPE shows is how much the future returns on equities should exceed what you’d garner from the long bond. It’s your reward for choosing the risk of stocks over the safety of Treasuries.
The bulls love the Excess CAPE. They advise investors to ignore the looming perils raised by the regular CAPE and its earnings yield, and rely on the Excess CAPE that makes everything look fine.
What the Excess CAPE is telling us
During the tech bubble, the only other time the CAPE was this high, the Excess CAPE was also pointing to disaster. That’s because “real rates,” as in most periods, were positive rather than negative, and in those times of strong economic growth, they were solidly in the plus zone. For example, in December 1998, the CAPE earnings yield was 2.58%, and the real Treasury rate was 1.53%, for a tiny Excess CAPE of 1.09%. Clearly, that thin margin of stocks over bonds wasn’t nearly enough for investors in the longer term, who dumped overpriced shares, driving down prices and driving up the Excess CAPE to the point where it provided a thicker cushion.
But today, the Excess CAPE doesn’t look nearly so dangerous. It appears reasonable and nonthreatening at the same time the regular CAPE sounds the alarm. Here’s why: Once again, the CAPE earnings yield is a lowly 2.60%, close to the number in late 1998. But the “real rate,” as most recently calculated by Shiller, is a negative 0.46%. That bond yields are lagging inflation increases the margin by which equities beat bonds, by a lot.
In most periods, you’d subtract a positive real rate from the CAPE yield and get an Excess CAPE lower than the regular CAPE, because Treasuries are yielding more than the rate of inflation. But not today. While the regular CAPE yield is 2.60%, the Excess CAPE is 3.06%. (That’s the 2.60% plus the 0.46% minus rate on Treasuries.) Even though the CAPE yield is extremely low, the Excess CAPE shows that compared with bonds, stocks are still pretty attractive. You can expect to reap 3.06% more from the U.S. big-caps going forward than from the long bond.
Keep in mind that a decent Excess CAPE of 3.06% doesn’t mean stocks will do well. The regular CAPE is still portending just 2.60% real returns—as we’ll see, at best. The Excess CAPE is really saying that Treasuries are a lousy choice, and stocks are just less lousy. But the bulls see another big virtue in the Excess CAPE. As long as low rates mean that stocks should far outperform bonds, what look like oversize P/Es are fully justified.
It’s true that stocks compete with bonds, and that lower real rates go hand in hand with richer valuations. P/Es could indeed be higher than their historic averages well into the future. But how high? They won’t stay at these levels for a simple reason: Real rates have never remained negative for long, and they won’t this time. Since the tech bubble burst, the U.S. economy has been growing much more slowly than in the 1980s and 1990s. That’s reduced demand for capital, lowering real yields.
That doesn’t mean negative rates are here to stay. From 2013 to 2018, inflation-adjusted yields on the 10-year toggled between 0.5% and 1.0%, and reached 1.52% in late 2018 when the Fed restored its easy money stance.
Rates will rise again, pounding stock returns
In its budget projections released in early July, the CBO forecast that the 10-year yield will rise from today’s 1.35% to 2.4% in 2024, and average 3.2% from 2026 to 2031. That outcome would see the return of real rates to at least 1% by the middle of the decade. They could go much higher as the Fed steps away from buying most newly issued government bonds. Then, our pool of domestic savings may not be sufficient to fund the huge future deficits, so that the volumes of newly issued bonds exceeds demand, pushing up real yields.
The coming increase in real rates will make the Excess CAPE look scary, just as the regular CAPE looks now. It’s also important to recognize that the CBO projects GDP growth of only around 1.5% from 2023 through the close of the decade. A sluggish economy will provide far less latitude for profit growth than in the past. Research Affiliates, a firm that designs investment strategies for $166 billion worth of ETFs and mutual funds, predicts that over the next decade, U.S. big-cap stocks will deliver nominal annual earnings growth of 3.5% and dividends of 1.5%. But because P/Es will fall and cause a big headwind for stock prices, investors will pocket returns of only 1.7% a year.
In other words, they probably won’t make enough to match the increase in their costs of buying groceries and paying rent. Falling rates provided a huge windfall for today’s investors. Their mistake is thinking that what has never lasted long will last this time. Going from what’s been billed as “the new normal” to the real normal will be a rocky road.
Subscribe to Fortune Daily to get essential business stories straight to your inbox each morning.