Latest Thoughts on U.S. Market Valuations

Professor Jeremy Siegel recently appeared on CNBC and suggested that the U.S. equity markets were approaching fair value and that markets might “pause” some of their strong gains in 2018. Siegel still believes corporate tax cuts are one factor that supports the market strength and that earnings should receive a boost from pending changes. He saw about 5% more gains in 2017 and predicts stocks having more difficulty next year.

Absent the corporate tax plan, we’d have a cloudier picture for U.S. stocks. But one question continues to come up in conversations: what about extended market valuations and the cyclically adjusted price-to-earnings (CAPE) ratio?

A piece on U.S. market valuations I posted earlier in the year has generated a lot of interest, and I updated it with more recent data below.

Robert Shiller and Jack Bogle are market prognosticators who tend to be more subdued in their outlook for U.S. equities. Earlier this year, Bogle presented his outlook for 10-year returns at the annual Chartered Financial Analyst conference in Philadelphia and suggested we’ve seen strong gains in the markets over the last 35 years that resulted from valuation expansion, and hence had a subdued outlook. Bogle’s model was fairly simple: take the 2% dividend yield on the market today, add in his personal estimates of 4% earnings growth and subtract 2% from speculative market activity or his anticipation of a decline in valuation ratios over the coming decade, and you come up with an outlook for 4% returns over the coming decade. If we assume there is 2% inflation, this would lead to just a 2% real return after inflation. Note that this is largely similar to Shiller’s outlook for returns from high CAPE ratios.

One chart that I think is not talked about enough in the context of valuation changes on the market is the dividend payout ratio of the market. I show a smoothed 10-year average dividend payout ratio in the spirit of Shiller’s 10-year smoothed earnings for the CAPE ratio. Prior to 2000, the dividend payout ratio averaged 60%. Since 2000, the dividend payout ratio has averaged 40%. This change in the nature of how firms reinvest their earnings, conduct stock buybacks and pay dividends is absolutely critical to the future earnings growth we are likely to get.

10-Year Average Smoothed Dividend Payout Ratio

Those who assume that earnings growth rates will revert to some historical average growth rate when firms paid out 60% of their earnings as dividends are assuming that all this money not being paid out—used for either buybacks or other reinvestment in business—is being completely wasted. That is an incorrect assumption, in my view.

This chart looks at the rolling 10-year and 20-year earnings growth rates of the CAPE earnings per share (EPS) that Shiller uses to make his dour forecasts on the market. If these numbers were to mean revert, that would be a cautionary tale for the markets. But in my view, the earlier declining dividend payout ratio means we are likely to see upside changes to these earnings figures. What is possible?

Growth Rates in 10- and 20-Year CAPE EPS

Changing dividend payout ratios have already translated to better earnings growth. Prior to 1982, the average dividend yield on the U.S. equity market was approximately 5% per Shiller’s data, and we had an average dividend payout rate of nearly two-thirds of earnings paid out as dividends. With only a third of earnings reinvested, firms were still able to achieve earnings growth of 3.3% per year.

Since 1982, payout ratios declined to an average of 51%, while at the same time firms started conducting stock buybacks. The average EPS growth during this period of reducing dividend payout ratios was an increase of 230 basis points (bps) per year, from the previous long-term average of 3.3% per year to 5.6% per year. In just the latest year, there has been a large increase in earnings, with a 20% year-over-year earnings growth as of the latest September quarter, with an average dividend yield of 2% and average dividend payout ratio of 46.7%.

When we look at the last 20 years, and particularly the last seven, we see consistent signs of 2% dividend yields with 2% net buyback ratios. The latest trailing 12-month net buyback for the S&P 500 was 1.84%. These net buybacks are going to continue to support earnings growth for the 10-year look-ahead period. These firms have locked in future EPS growth because they reduced their shares outstanding.

Returning to the table above, where I showed the earnings growth since 1982 as being higher than the previous 110 years, the current dividend payout ratios are consistent with an even further drop in the payout ratios from their average since 1982. I can see a case that earnings growth picks up even from that 5.6%-per-year mark that we had for the period 1982–2017. It would not surprise me to see earnings growth of 6% to 7% per year over the next decade.

The standard pushback is that firms are just leveraging up to conduct buybacks—that interest rates are at historical lows, leading to higher margins than are sustainable. The reverse case is that the changing composition of companies—into higher-margin businesses that have more revenue abroad with lower tax rates than in the U.S.—also means margins may not be mean reverting anytime soon either. Of course, no one knows how the future will unfold, including me.

The charts above caution anyone relying on historical patterns of earnings growth trends from over-extrapolating them into the future. Professor Siegel looks at the current earnings yield of the market associated with a 20 price-to-earnings ratio and thinks 5% is a pretty good indicator of long-term, after-inflation real returns. Add in inflation of 2% and you get 7% nominal returns. This is a touch below their historical 6.5% to 7% that he showed in Stocks for the Long Run as being the historical return to U.S. equities, but it is not dramatically different. I think his model for looking at the markets makes more sense than some of these more dour predictions—for what that’s worth.