Stock prices fell sharply before the Labor Day weekend, following an astonishing rise from March lows. Information technology stocks in the S&P 500 fell from a peak of 22128.1 on September 2 to 2046.27 by 2:30 two days later. But that was just the equivalent of losing two weeks of gains.
Momentum trading had surely pushed some high fliers into frothy territory, so a correction is correct. But skeptics and short traders claim the market as a whole still remains totally disconnected from “fundamentals.”
Other commentators suggested investors were suddenly terrified that Congress had not yet agreed to borrow and give away trillions more to “stimulate” something or other. “If we do not get another stimulus aid package in time,” warns Forbes columnist Naeem Aslam, “the economic recovery will remain fragile”(though it’s growing at a 26.2% to 29.6% rate). A recent Wall Street Journal story likewise warned that “failure by Congress to deliver additional relief measures for American consumers and businesses could weigh on market sentiment.”
Yet the disappearance of the “stimulus package” is old news. Most of it began and ended in April when nearly everyone received a $1,200 check, and PPP loans soon dried up. The extra $600 of weekly unemployment benefits ended July 31. Yet stocks were particularly strong in August. The rapid winding down of stimulus schemes from May to July was essentially irrelevant compared to April’s all-important news that 28 states announced that they would be reopening from April 20 to May 4, soon followed by most others.
This brings us back to the more serious question – namely, the alleged disconnect between S&P 500 stock prices and “fundamentals.” This often boils down to the fact that stock prices seem high relative to past earnings. But the price/earnings (P/E) ratio can’t be understood without looking at bond yields. That is because a lower interest rate increases the discounted present value of future earnings.
The argument for stocks being greatly “overvalued” rests on the fact that the trailing P/E ratio rose significantly from May 1 to September 1. On January 1, the P/E ratio was 24.21 –about the same as two years before (24.87). Even after Covid-19 and lockdowns crushed the economy, the P/E ratio was still 23.74 on April 1. Stock prices and earnings had both collapsed in sync. The P/E ratio was 25.10 on May 1 after the Fed funds rate fell to nil and the $1,200 checks and PPP loans peaked. It then rose to 26.69 on June 1, 27.57 on July 1, 28.31 on August 1 and 30.32 on September 1.
Using the first of the month as a rough gauge of trends, S&P 500 stock prices have been rising even faster than these firms’ average earnings over the preceding 12 months. This may appear to suggest that stocks grew overvalued relative to the growth of earnings, but there are two reasons to question that conclusion. The first reason is that we need to compare these historically high P/E ratios with historically low bond yields. The second is that we need to realize that trailing earnings over the past 12 months currently translate into a ridiculously low denominator for the P/E ratio. Average earnings over the past 12 months include some of the worst in U.S. history – notably, March and April.
To compare both P/E multiples and bond yields over time it is illuminating to simply invert the P/E ratio in order to create an E/P ratio. This E/P ratio is called the “earnings yield.” The earnings yield is simply the inverse of the trailing price-earnings (P/E) ratio. That is, the earnings yield shows earnings per share for the last twelve months divided by the current market price per share.
Ed Yardeni depicts the earnings yield as a blue line in the graph reproduced here (his Figure 8). The 10-year Treasury bond yield is shown in red. Obviously they generally rise and fall together, sometimes with the earnings ratio in the lead.
In 1997, Yardeni dubbed the relationship between bond and earnings yields “The Fed Curve.” That was because the Greenspan Fed had just alluded to it. Later, I modestly renamed it “The Reynolds Model” because I unveiled it in a March 21, 1991 memo to my consulting firm’s institutional investors. Alan Greenspan might have seen it, since he was on my comp list after because we worked together on President Reagan’s transition team.
Since the earnings yield is the P/E ratio inverted, the graph shows that the P/E ratio falls (the blue line rises) when inflation speeds up, particularly the decade after the gold window closed on August 16, 1971. Conversely, the earnings yield falls and the P/E ratio rises when inflation and bond yields fall. If the Fed were to succeed in getting inflation significantly above 2%, they would also succeed in pushing the P/E ratio down. Chairman Powell announcing the Fed’s hope for higher inflation on August 27 was risky news for stocks, and for bonds. The 10-year bond yield rose from 0.65% on August 24 to 0.72% on September 4, which may partly explain the observed decline in the P/E ratio.
What happened to stock prices in recent months is quite consistent with the Yardeni/Reynolds Model: The higher P/E ratio (lower E/P ratio) can be explained by the 10-year bond yield dropping from 1.88% on Jan 2 to 1.1% by March 2 and then 0.62% on April 1 before rising only very recently to 0.72%. The Fed funds rate deserves no credit for lower bond yields (worldwide), because the Fed reluctantly followed the market down – keeping the fed funds rate at 0.6% in March while the 3-month bill rate fell to 0.29%.
The Federal Reserve can certainly crash the market (e.g., the double-dip recessions of 1980-82), but milder forms of Fed activism rarely explain bond yields or stock prices. The 10-year bond yield has at times risen 3.5 to 4 percentage points above the Fed funds rate, as in 1992, 2001-04 and 2010. Also, the S&P 500 stock index hovered at or below 2000 the last time the Fed kept the funds rate near zero in 2014-15, then rose 46% by July 2019 even as the Fed raised the funds target ten times to 2.5%.
Yardeni’s graph leaves no room for fiscal stimulus as an explanation for the stock market’s perfectly predictable rebound after State lockdowns began to be relaxed. Earnings rose because closed businesses were permitted to open. And the P/E ratio rose because the earnings yield always falls when bond yields fall.
Some see today’s high P/E as prima facie proof of a speculative bubble. But the fact that stock prices may look high relative to earnings over the past 12 months (which include an unprecedented GDP contraction) does not mean they are high relative to the next 12 months.
For those who find stock price gains inexplicable and therefore terrifying, the Yardeni-Reynolds Model offers a purely empirical explanation of a seemingly high trailing P/E ratio – extremely low global inflation and bond yields combined with an artificially low P/E denominator of past earnings that includes the terrible second quarter of 2020.