The stock market looks like a reasonable value if you assume earnings are stable at this high level. Yet, earnings are not stable and the U.S. stock market (SPY), (IVV) faces great risk. In the past 12 years, earnings have been the most volatile in history and the next big plunge may have already started.
With 99.2% of the 500 companies reporting, "as reported" earnings for second quarter are $21.62, down from $23.03 in the first quarter. Three months ago the official S&P estimate for the second quarter was $26.09, but has fallen fairly steadily as actual earnings were reported. With this decline, 12-month earnings declined from $88.54 last quarter to $87.93, although they are still up slightly year over year.
Here are the earnings, adjusted for inflation:
Click to enlarge images.
A stark pattern of increased earnings volatility emerges if you look at the annual change in real earnings (black line in chart below) compared to annual GDP growth (red line in chart below).
From the 1950s to the mid-1980s, earnings fluctuated about 3.8 times more than GDP. Since then, earnings growth has become increasingly volatile and moves 30 times or maybe a 100 times more than economic growth. The decline of earnings starting in 2000 was the sharpest since the Great Depression. Every turn in earnings growth since has been more volatile than the one before. Earnings spiked 73% in 2003 and collapsed 89% in the Great Recession, dwarfing any previous decline. The surge in growth that peaked in the second quarter of 2010 is off the chart seven times over. The growth rate in earnings has fallen steadily for two years. The coming decline may be just as dramatic.
Part of the earnings volatility stems from the incentive performance pay given to CEOs. Options often have to be held three years. If a CEO can generate three years of stellar growth, he or she hits the pay bonanza. It's easier to hit the pay bonanza if you start from a low base. A decline in the next year or two would set up a low base for the next run at the big payoff.
The chart suggests the economy needs to grow about 2.5% for earnings to hold steady. Stronger growth means the highly leveraged earnings can soar rapidly while weaker growth means the leverage works in reverse and earnings sink rapidly, setting up a low base. The current GDP growth rate implies earnings growth could soon turn negative.
It looks like earnings have run as far as they can. Earnings are about 38% above the long-term growth rate suggested in the first chart. This 1.9% real growth rate is actually down from a 2.1% long-term growth rate in 2000. Earnings growth in the last decade has both slowed and become more volatile. We are probably about to see how nasty this combination can be.
Looking at a P/E based on the past 12 months' earnings, or estimated future earnings, is a sucker's game when earnings end a volatile move up. A volatile rise in earnings makes the market appear less overvalued than it really is. Estimated earnings exaggerate the effect. As earnings began plunges in 2000 and 2007, estimated future earnings were projecting double-digit growth, as they are now.
With earnings growth at the most volatile level in history, looking at earnings over a longer period than 12 months becomes more important than ever. This is where Shiller's PE10 or my PEses show their value. For more information on PEses, see my article titled "The Stock Market's Worst 17 Years."
With August's rise to 36.7, the PEses is a fraction higher than the peak valuation in 1929. Unlike a P/E based on the past 12 months, the PEses has a strong correlation with future return. Here is the correlation with 17-year periods since 1960. The chart aligns the PEses at the beginning of a 17-year period with the annualized return after inflation and dividends during that 17-year period:
From this high level of valuation, the chart suggests the return on the S&P 500 over the next 17 years will be about 0% and that we are 12 years and nine months into the worst 17-year period the stock market will ever have. While the PEses is the best measure of value I have found, it is a poor timing tool. While I suspect the market will decline 70% in the next one to four years, I have certainly mistimed the start of the decline.
The potential downside in the market is likely 15 times the potential upside in the next four years.
Disclaimer: There is no guarantee analysis of historical data and trends enable accurate forecasts. The data presented is from sources believed to be reliable, but its accuracy cannot be guaranteed. Past performance does not indicate future results. This is not a recommendation to buy or sell specific securities. This is not an offer to manage money.
