Yardeni on S&P 500 Earnings, Valuation, and the Pandemic: Part II

This three-part series is excerpted from my 2020 book S&P 500 Earnings, Valuation, and the Pandemic (co-authored with Joseph Abbott). Part I (Earnings): Introduction. Discounting Forward Earnings. Lots of S&P 500 Earnings Measures. Revenues, Earnings & Profit Margins. Part II (Valuation): Flying with the Blue Angels. In the Eyes of the Beholder. Reversion to the Mean. Fundamentals Matter. Discipline of Dividends. Part III (Pandemic): Very Useful Indicators. The GFC and the GVC. Fed-Led Valuation Meltup. Epilogue. 

Part II: Valuation

Flying with the Blue Angels

Industry analysts provide the earnings estimates that are discounted in stock prices. Investors determine the valuation of those earnings. While forward earnings isn’t an infallible measure of earnings for forecasting purposes, we are convinced that the market is discounting the time-weighted average of analysts’ consensus earnings expectations for the current year and the coming year. In our analysis of the stock market equation, we can specify that “E” is S&P 500 forward earning per share. “P” is the S&P 500 stock price index. “P/E” is the ratio of the S&P 500 stock price index to the S&P 500 forward earnings per share. Industry analysts’ consensus expectations are used to derive the forward E, and investors determine the forward P/E.

We devised our “Blue Angels” chart framework to monitor these three variables in a visually useful way. In the monthly version, we multiply the S&P 500’s forward earnings per share by hypothetical forward P/Es of 5.0 to 25.0 in increments of 5.0 (Fig. 21). The result is five different time series of an implied S&P 500 index price at the various P/E levels. They move in a parallel formation and never collide, just like the vapor trails left behind the Navy’s famous Blue Angels jets.[1]

Superimposing the actual S&P 500 stock price index shows when it is breaking into new valuation territory, i.e., changing P/E “altitude” by moving toward a new higher or lower P/E series trail. In our presentations to clients, we often refer to the S&P 500 series as “the stunt plane flying through the vapor trails of the Blue Angels.” The same framework can be constructed using weekly data to keep a more frequent watch on the Blue Angels relationships among P, E, and P/E (Fig. 22).

As the S&P 500 ascends or descends through the Blue Angels indexes, we can see how much of the move is attributable to forward earnings versus the forward P/E. Generally, S&P 500 forward earnings per share isn’t as volatile as the index’s forward P/E. So big short-term moves in the stock index most often reflect changes in the forward P/E that cause the index to climb or fall toward the next Blue Angel P/E vapor trail. Conversely, moves in the actual index price that do not bring it closer to a nearby Blue Angel P/E altitude confirm that changes in the earnings outlook are driving the S&P 500’s price action.

Forward earnings per share tends to rise fastest during economic recoveries and to fall fastest during recessions. Our simple Blue Angels framework clearly shows that bull markets typically occur when forward earnings and forward valuation are rising. Bear markets, when the S&P 500 is down 20% or more, typically occur when both are falling. Short-term bull-market selloffs of 10% to less than 20%, a.k.a. “corrections,” occur when valuation declines while forward earnings continues to rise. There was a rare bear market in 1987 when the forward P/E fell sharply while forward earnings continued to rise.

The bottom line is that we use Earnings Squiggles and Blue Angels as tools to benchmark our own forecasts to the earnings expectations and the valuation levels that the market is discounting. We monitor the trends of earnings expectations for the current year and next year as well as for forward earnings. We watch to see how much altitude the S&P 500 “stunt plane” is gaining or losing as a result of changes in forward earnings and valuation. These tools are like our air traffic control system, providing us with guidance through market turbulence.

So, for example, in the bull market from March 2009 to February 2020, as in the previous one from 2003 through 2007, we stayed bullish during stock price swoons when we saw that forward earnings was still rising. We anticipated a correction in early 2020 because the valuation multiple seemed high to us when it rose to 19.0, matching the P/E just before the big correction in late 2018.[2] We turned bullish on March 25, 2020, anticipating that the Fed’s latest round of monetary easing would boost the P/E with stock prices rising faster than earnings were falling, as discussed in Part III.

So, what we actually do for a living is this: we forecast the forecast. We predict where forward earnings will be at the end of the current year and the coming year. That amounts to forecasting next year’s and the following year’s earnings since those will be the forward earnings at the end of the current year and the coming year, assuming that industry analysts eventually will concur with our earnings outlook. Of course, it doesn’t end there. To convert our forecasts for forward earnings per share to S&P 500 targets at the end of the current year and the coming year, we also need to forecast where the forward P/E will be at both points in time.

In the Eyes of the Beholder

As an economist, I’ve always felt relatively comfortable with predicting earnings, since they are mostly determined by the performance of the economy. Assessing the outlook for the P/E is the tougher of the two variables to forecast, in my opinion.

Judging valuation in the stock market is akin to judging a beauty contest. Episode 42 of the television series The Twilight Zone is titled “Eye of the Beholder.” It’s about a woman who undergoes her 11th and last legally allowed facelift to correct her looks, as required by the totalitarian regime. When the bandage is removed, the doctors are disappointed and can barely hide their disgust: she is still beautiful. Then the camera reveals the faces of the doctors and nurses. They look horrifying to us viewers, with their pig-like snouts, though clearly pleasing to one another. The beautiful misfit escapes with a handsome man to a village of their “own kind,” where the rest of society won’t be subjected to their repellent good looks. “Beauty is in the eye of the beholder,” the man tells the woman.

Not only is beauty subjective, Hollywood tells us, but it can be dangerous. At the end of the original version of the movie King Kong (1933), the big ape’s death is blamed by his handler on Ann Darrow, Kong’s blond love interest, played by Fay Wray: “It was beauty that killed the beast.”

Valuation is in the eye of the beholder too. And buying stocks when they are most loved and very highly valued can also be deadly. For example, during the late 1990s, investors scrambled to purchase high-tech stocks in the United States. At the height of the frenzy, the S&P 500 Information Technology sector accounted for a record 33.7% of the market capitalization of the entire S&P 500 but only 18.2% of its earnings (Fig. 23). The forward P/E of the S&P 500 peaked at a record 24.5 during July 1999, led by a surge in the forward P/E of the S&P 500 Information Technology sector to a record high of 48.3 during March 2000 (Fig. 24). When that bubble burst, many investors suffered crushing losses in their portfolios.

Stocks tend to rise along a long-term trend line that is determined by the long-term growth rate of earnings. Since 1979, the trend growth rate for S&P 500 forward earnings on a monthly basis has ranged between 6% and 7% (Fig. 25). Nevertheless, a long-term investor who hopes to earn this projected return may earn less if stocks are bought when they are overvalued and earn more if stocks are bought when undervalued. A short-term trader doesn’t care about long-term returns, but buy-and-hold investors should care about buying stocks when they are relatively cheap rather than too expensive.

How can we judge whether stock prices are too high, too low, or just right? Investment strategists are fond of using stock valuation models to do so. Some of these are simple. Some are complex. The levels, changes, and growth rates in numerous variables—such as earnings, dividends, inflation, interest rates, and various risk metrics—all are thrown into the pot to cook up a “fair value” for the stock market. If the stock market’s price index exceeds the number indicated by the model, then the market is overvalued. If it is below fair value, then stocks are undervalued. As a rule, investors should buy when stocks are undervalued and should sell, or hold off buying, when they are overvalued.

A model can help us to assess value. But models by their very nature are attempts to simplify reality, which is always a great deal more complex and unpredictable. Valuation is ultimately a judgment call. It tends to be controversial, since everyone has their own opinion on what’s a pig of a stock and what’s a knockout at various levels of valuation.

Valuation is not only subjective; it’s also relative.

Stocks are cheap or dear relative to other assets, such as bonds, for example. There are no absolutes. Even this statement is controversial since some observers swear by a reversion-to-the-mean approach, which compares stock valuation to its historical average rather than to other asset classes. When the P/E is above the historical mean, they warn that stocks are overvalued and vulnerable to reverting to the mean. 

These diehards ignore all other factors that may be boosting valuations, and occasionally die waiting to be proven that they were right after all. It’s been said that history doesn’t repeat itself, but it rhymes. Similarly, history shows that valuation multiples do eventually revert to their means, though only briefly as they transition from overvalued to undervalued and back. Insights into how much time they might linger above and below their means and the magnitude of the deviations are not provided by simple reversion-to-the-mean models, which also don’t consider that means can change over time along with inflation and interest rates.

Reversion to the Mean

Of the various reversion-to-the-mean models, it’s the deviation of the forward P/E from its mean that we weight the most when we assess valuation. But the exercise is still a beauty contest. Common sense strongly suggests that the best time to buy stocks is when forward P/Es are low, while the best time to sell is when P/Es are high. However, doing so is not that simple. Stocks seemed relatively expensive in late 1996, which is why Federal Reserve Board Chair Alan Greenspan famously asked the valuation question in a December 5, 1996 speech: “But how do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions, as they have in Japan over the past decade?”

Right before posing the question, he suggested that stocks were not irrationally exuberant given that “sustained low inflation implies less uncertainty about the future, and lower risk premiums imply higher prices of stocks and other earning assets.” The S&P 500 proceeded to soar 106.5% for another three years, from December 6, 1996 through March 24, 2000, led by a bubble in technology stocks. Lots of money can be made during bubbles, if you know enough to get out at or near the top. 

Another issue regarding forward P/E data for the S&P 500 is that it’s available only from September 1978 onward. More years of data are necessary to determine whether the valuation multiple is high or low within an historical context. Besides, more data might suggest other testable models of valuation. As noted in Part I, there are other earnings series of actual results—on both reported and operating bases and going back much further than forward earnings—that can be used to construct other reversion-to-the-mean models. Often, these models are based on four-quarter moving sums of the earnings series. In other words, “trailing earnings” are used to calculate “trailing P/Es.”

The advocates of trailing earnings models do have the choice of using either reported earnings or operating earnings, i.e., excluding one-time extraordinary gains and losses. Of course, more pessimistically inclined investment strategists focus on reported earnings, the lower of the two measures. Whichever is used, the data are available only on a quarterly basis with a lag of three to six weeks, limiting the usefulness of a trailing-earnings approach. In any event, stock prices should be based on expected earnings, not trailing earnings, in our opinion. Forward earnings data reflect expectations and are available on a timelier basis, though with less history than trailing earnings. Models with P/Es based on trailing earnings often produce valuation conclusions quite different from those of models with P/Es based on forward expected earnings.

Using monthly data dating back to 1989, let’s compare the valuation multiples of the S&P 500 when using forward earnings, trailing operating earnings, and trailing reported earnings. The P/Es based on trailing earnings—both operating and reported—always exceed the measure based on forward earnings. And the P/E based on trailing reported earnings always exceeds the P/E based on trailing operating earnings (Fig. 26). Given that they are using past earnings, trailing P/Es tend to imply overvaluation well ahead of forward P/Es, so investment strategists relying on trailing earnings tend to turn bearish too early in bull markets. The most bearishly inclined of them tend to favor the trailing P/E based on reported earnings because it is the most pessimistic of the bunch.

To be fair, when recessions hit, forward earnings expectations turn out to be too high and are slashed, confirming with the benefit of hindsight that forward P/Es were too high. The bears tend to growl, “We told you so.”

Joe and I do track all the measures of the P/E both in absolute terms and relative to their means, along with similar valuation ratios deemed to be mean reverting. However, we don’t buy the idea that the mean is determined by the laws of nature and exerts some sort of inherent gravitational pull on valuation, surrounding it with a force field that deflects all other influences. While the simple reversion-to-the-mean models are worth tracking, in our view, we recognize that they ignore how changes in interest rates, inflation, and technologies might impact valuation on short-term and long-term bases.

In any event, we’ve constructed a P/E series that starts in 1935 using S&P 500 quarterly trailing reported earnings through 1978, monthly forward earnings from January 1979 through April 1994, then weekly forward earnings (Fig. 27). The mean of this patchwork has been around 15.0. The series shows that its usefulness as a market-timing tool leaves much to be desired. It can take a long time to revert to the mean both on the way up and on the way down. On the other hand, this P/E series did show that stocks were cheap relative to the mean in the early 1980s, expensive in the late 1990s, and cheap again during the late 2000s.

Another valuation measure favored by the reversion-to-the-mean crowd is the ratio of the value of all stocks traded in the US to nominal GNP, which is nominal GDP plus net income receipts from the rest of the world. The data for the numerator are included in the Fed’s quarterly Financial Accounts of the United States and lags the GNP report, which is available on a preliminary basis a couple of weeks after the end of a quarter. They aren’t exactly timely data.

This ratio has been widely followed ever since Warren Buffett highlighted it in an essay for the December 2001 Fortune: “For me, the message of that chart is this: If the percentage relationship falls to the 70% or 80% area, buying stocks is likely to work very well for you. If the ratio approaches 200%—as it did in 1999 and a part of 2000—you are playing with fire.”[3]

We can construct both monthly and weekly proxies for the Buffett Ratio. The S&P 500 stock price index can be divided by S&P 500 forward revenues per share instead of forward earnings per share. This forward price-to-sales ratio (P/S) closely tracks the Buffett Ratio (Fig. 28 and Fig. 29). However, this forward P/S ratio is very highly correlated with the forward P/E ratio, so it doesn’t really bring much additional value to assessing valuation (Fig. 30). And neither does the Buffett Ratio for that matter.

Fundamentals Matter

A closer look at our P/E series since 1935 shows that the mean since then doesn’t mean much, since inflation and interest rates likely influenced the valuation multiple. The P/E was generally below the historical average when inflation and interest rates were rising toward historically high levels. It was generally above the average when inflation and interest rates were falling toward historically low levels (Fig. 31 and Fig. 32). These fundamental factors obviously matter in the determination of valuation. Valuation isn’t all about reversion to the mean!

Many years ago, from the late 1970s through the late 1990s, there was a reasonably good correlation between the 10-year US Treasury bond yield and the S&P 500 forward earnings yield, which is simply the reciprocal of the forward P/E (Fig. 33). In 1997, I called this relationship the “Fed’s Stock Valuation Model” (FSVM), and the name stuck. I must have cursed it, since it hasn’t worked as a useful valuation model or market-timing tool since the early 2000s. The FSVM has been signaling that stocks are increasingly cheap relative to bonds since the early 2000s (Fig. 34). The FSVM certainly didn’t provide any warning ahead of the grizzly bear market caused by the Great Financial Crisis.

Maybe it is starting to work again now.

The model showed that stocks were undervalued relative to bonds by a record 88% during the week of August 7, 2020. After all, the Treasury bond’s record-low yield of 0.50% on August 4 implied a P/E for the bond of 200 (!), using the reciprocal of the yield. In other words, the FSVM was clearly signaling the S&P 500’s forward P/E was too low relative to the bond’s P/E. Take that for what it’s worth considering that it is just as easy to argue that bonds were ridiculously overvalued relative to stocks. The truth was somewhere in between, i.e., stocks were relatively cheap, while bonds were relatively expensive.

Now let’s examine the impact of inflation on valuation. The earnings yield of the S&P 500, which is simply the reciprocal of the P/E based on reported earnings, is well correlated with the CPI inflation rate on a year-over-year basis (Fig. 35). The real earnings yield (REY) of the S&P 500 is the difference between the nominal yield and the inflation rate (Fig. 36). The result is a mean-reversion valuation model that logically includes inflation.

The average of the real yield since 1952 is 3.19%. The model tends to anticipate bear markets when the yield falls close to zero. John Apruzzese, the Chief Investment Officer of Evercore Wealth Management, examined this model in his November 2017 paper, A Reality Check for Stock Valuations. Based on the REY model, he found that “stocks appear more reasonably priced than the conventional P/E ratio suggests during periods of low inflation and rising markets, and more expensive during periods of high inflation and falling markets when they otherwise might seem cheap.”

Another fundamental factor that needs to be considered is the analysts’ consensus expected long-term earnings growth rate (LTEG) for the S&P 500. I/B/E/S provides a series for LTEG, i.e., average projected annual earnings growth over the next five years (Fig. 37). It’s available on a monthly basis since 1995 and on a weekly basis since 2006. It has been quite volatile considering that it is supposed to measure consensus expectations for the long-term trend in earnings growth. In Part I, we observed that both reported quarterly earnings since 1935 and forward operating earnings since 1979 have ranged between 5% and 7% annualized growth trends.

So how can we explain why the monthly measure of LTEG has ranged between a low of 9.3% and a high of 18.7% with a mean of 12.6% from 1995 through the end of 2019? Keep in mind that LTEG is based on analysts’ expectations for the long-term earnings growth of the companies they follow, not for the overall S&P 500. Their optimistic bias toward the future of their companies is clearly reflected in LTEG.

In addition, as was evident during the second half of the 1990s, industry analysts justified the run-up in tech stock prices by raising their LTEG expectations from 16.7% in January 1995 to a record high of 28.7% during October 2000 (Fig. 38). When the tech bubble burst, LTEG reversed course. It was pushed further downward again by the Great Financial Crisis. It was back on the ascent from 2017 through the first half of 2018, when industry analysts anticipated that President Trump’s pro-business policies—including less regulation and lower corporate tax rates—would be bullish for long-term earnings growth. But then during the second half of 2018 through 2019, Trump’s escalating trade wars caused LTEG to tumble, and the coup de grace was provided the following year by the Great Virus Crisis.

We can use the monthly and weekly LTEG series to construct a PEG (or P/E to long-term earnings growth) ratio for the S&P 500. It is equal to the forward P/E divided by LTEG (Fig. 39 and Fig. 40). Conceptually, using a PEG ratio for valuation purposes makes lots of sense. When long-term investors buy stocks, they aren’t focusing on expected earnings just over the coming year but rather over the next several years. The higher the expected growth in earnings, the more dearly an investor is likely to value a stock.

However, what makes sense for an individual stock may not be as sensible for the overall market.

Overall earnings growth is limited by the nominal growth rate of overall revenues, which depends on the growth of nominal global economic activity. If earnings growth expectations for the overall market well exceed the growth potential of the global economy, that would be a sign of irrational exuberance rather than justification for paying an inflated P/E well above an inflated LTEG. So while tech P/Es seemed to be justified by rising LTEG for tech stocks during the late 1990s, they had more room to crash when LTEG was revised downwards when the bubble burst.

In other words, the PEG measure is more likely to run into trouble at the sector level, where irrational exuberance may distort earnings growth expectations—as happened during the second half of the 1990s, when industry analysts raised their LTEG for technology. It seems to us that they were doing so mostly to justify the rapid ascent in prices. Even Fed Chairman Alan Greenspan justified high stock prices during the tech bull market by noting that industry analysts were raising their growth expectations. He said so in a September 5, 1997 speech at Stanford University: “And the equity market itself has been the subject of analysis as we attempt to assess the implications for financial and economic stability of the extraordinary rise in equity prices—a rise based apparently on continuing upward revisions in estimates of our corporations’ already robust long-term earning prospects.”

What he forgot to mention is that this is exactly why irrational exuberance always ends badly. It attracts too much buzz, too much press, too many speculators, and most importantly too much capital that funds new competitors in the hot industry. While valuations are soaring, competition from new entrants starts to squeeze margins and saturates the market. Once investors recognize that analysts’ earnings growth expectations are too optimistic, stockholders scramble to sell, causing P/Es to fall even faster than growth expectations.

Discipline of Dividends

The focus on valuing earnings is a relatively new phenomenon that started with the bull market of the 1990s. Before then, most valuation models for individual stocks focused on dividends, not earnings. Investors compared the dividend yield, not the current earnings yield, to the bond yield. Corporations were valued on their ability to pay and grow dividends, which represented a tangible return to investors. Retained earnings—profits after taxes and dividends—were reinvested in the business, presumably to increase the capacity of the corporation to pay more dividends in the future.

Investors could analyze the dividend payout history of a company. Then they could project a reasonable future payment stream to shareholders and calculate the present discounted value of the firm using a dividend-discount model (DDM). If the present discounted value was more than the share price in the market, then investors could expect to get a better-than-average return by investing in the company’s stock. Of course, if the present discounted value of the projected future stream of dividends was less than the share price in the market, then prudent investors might sell the stock, or at least underweight it in their portfolios.

The simplest version of the DDM is the Gordon growth model, which assumes a constant rate of growth for a company’s dividends (g). The fair value price (P) is calculated as the estimated value of next year’s dividend (D) divided by (r – g), where r is the company’s cost of capital:

P = D / (r-g)

This dividend-centric valuation discipline provided a powerful and conservative system of checks and balances for corporate managers. Dividends are cash payments. There is no way to print the money; it must be available from a company’s cash flow. Managers were under pressure to deliver dividend growth, but they also had to retain enough of their earnings to reinvest in their companies so that dividends would continue to grow.

This conservative but disciplined system was replaced during the bull market of the 1990s by a more freewheeling approach to valuation that was more easily abused to boost stock prices to levels that could never be justified by dividends. Indeed, many companies, especially those that seemed to be growing rapidly, reduced their dividend payouts or just eliminated them entirely. More earnings were retained, and fewer dividends were paid out to investors.

The rationale in most cases seemed sensible: growth companies experiencing rapid increases in their earnings could reinvest their profits and get a better long-term return for their investors through share price appreciation as the companies expanded and became more valuable. Besides, investors couldn’t reinvest the dividends on their own and still do as well as a growth company because they would have to pay taxes on the dividend income after it had already been taxed at the corporate level.

The new approach to valuation based on earnings rather than dividends was both a blessing and a curse. Under the dividend regime, most managers adopted a slow but steady and conservative approach. As long as dividends were growing, investors tended to be content with managements’ performance. Growing dividends was a long-term process occasionally disrupted by economic downturns. Reinvesting retained earnings required a great deal of planning, and the projected returns had to be just high enough to boost dividends without subjecting the company to a great deal of risk.

Under the earnings-centric valuation regime, companies no longer faced the quarterly grind of delivering cash dividends. The cash could be plowed back into the business. Greater risk was acceptable because there was less pressure to deliver the cash to investors every quarter. This meant that managers could be more entrepreneurial. It also meant that some could abuse the system by artificially boosting their earnings. Managing earnings rather than managing the business became an increasing problem during the bull market of the 1990s.

Then-Fed Chair Alan Greenspan put the stock market bubble of the late 1990s into perspective on March 26, 2002 in a speech on “Corporate Governance” that he presented at New York University. Greenspan observed that shareholders’ obsession with earnings was a relatively new phenomenon:

"Prior to the past several decades, earnings forecasts were not nearly so important a factor in assessing the value of corporations. In fact, I do not recall price-to-earnings ratios as a prominent statistic in the 1950s. Instead, investors tended to value stocks on the basis of their dividend yields."

During 2002, many of the most abusive practices in managing earnings came to light as a consequence of numerous corporate accounting scandals. Undoubtedly, the system needed to be reformed, and it was through the Sarbanes–Oxley Act of 2002. The good news is that dividends have been making a comeback since the financial crisis of 2008. As bond yields have plunged since then, dividend-yielding stocks have come back into favor, especially those of companies with a long record of raising their dividends. We often show the power of dividends by calculating the current dividend yield of an S&P 500 portfolio purchased in 1970, 1980, 1990, 2000, and 2010. By mid-2020, those hypothetical portfolios were yielding 69.2%, 44.0%, 18.0%, 4.5%, and 4.7%, respectively.

Unlike the forward earnings yield, or even the trailing earnings yield, the dividend yield of the S&P 500 doesn’t correlate enough with either the 10-year US Treasury bond yield or the inflation rate to inspire the construction of a valuation model (Fig. 41). As inflation soared from the 1950s through the 1970s, the bond yield rose much faster than the dividend yield. During the disinflationary 1980s and 1990s, the bond yield fell faster than the dividend yield. Since 2000, the dividend yield has been on a slight uptrend, while the bond yield has continued to fall. Since the financial crisis of 2008, the bond yield and the dividend yield have been about the same for the first time since the late 1950s.

We’ve extended our Blue Angels framework to track the relationship of the S&P 500 dividend and dividend yield to the S&P 500 stock price index (Fig. 42). The arithmetic relationship is simple:

P = D / Y

where P = the S&P 500 stock price index, D = aggregate dividends paid by the S&P 500 corporations, and Y = the dividend yield, or D/P.

The Blue Angels show the hypothetical value of the S&P 500 using the actual dividends paid out divided by dividend yields from 1.0% to 6.0%. The Blue Angels analysis reveals what’s propelling the S&P 500’s price performance. If the “plane” (index) is flying along a particular “vapor trail,” then the flight path of dividend is determining the flight path of the plane (i.e., the S&P 500’s stock price performance). Diversions above (below) a vapor trail indicate rising (falling) valuation since investors are clearly paying more (less) for the same amount of dividend dollars thus getting a lower yield.

As of this writing, in September 2020, the conclusion of this Blue Angels analysis is that stocks are attractive relative to bonds because the dividend yield exceeds the bond yield, and the long-term uptrend in S&P 500 dividends has been roughly 6% since the end of 1946 (Fig. 43).

___________________________

[1] The different time series lines in our Blue Angels charts would collide if forward earnings turned negative, but this never has happened for the broad market averages we track.

[2] In our January 28, 2020 Morning Briefing, we wrote that a P/E-led meltup increased the risk of a correction. We noted that the stock market started the year with “nothing to fear but fear itself.” By the end of January, the possibility of a pandemic gave us all something to fear.

[3] “Warren Buffett on the Stock Market,” Fortune, December 10, 2001.

___________________________

Institutional investors may sign-up for a complimentary trial of Dr. Ed's research service at https://lnkd.in/eYNwch8. Books and studies in his "Predicting the Markets" series can be found on his Amazon author's page.

Sign up to follow his LinkedIn Newsletter at https://lnkd.in/eyW46Dn

Don't miss each edition of Dr. Ed's Newsletter: Click Subscribe above or below to be notified.  

___________________________

Copyright ? 2020 Edward Yardeni. All rights reserved. No part of this publication may be reproduced in any form or by any electronic or mechanical means, including information storage and retrieval systems, without permission in writing from the publisher, except by reviewers, who may quote brief passages in a review. ISBN: 978-1-948025-08-9 (paperback) ISBN: 978-1-948025-09-6 (eBook)

Part I: Earnings .......... Part II: Valuation .......... Part III: Pandemic