Dear Clive,
OK, now we are getting to the guts of our argument about why the Dow could rise quickly to 36,000. Some of this stuff is complicated and arcane, but I’ll try to make it as clear as I can.
You say that, at a price-to-earnings ratio of 25, the stock market is yielding 4 percent–thus making it “fairly valued, not vastly undervalued,” as my colleague Kevin Hassett and I have suggested.
In this context, you cited Jeremy Siegel of Wharton, who criticized our March 30 piece in the Wall Street Journal. Siegel claimed our analysis had a “serious flaw”–and you join him in that conclusion.
Sorry, but you are both wrong. You assume that the real rate of dividend growth (or earnings growth) is likely very low, perhaps zero, because, as you say, “earnings rise with inflation.”
Ah, but that’s not right.
Let’s step back for a second. We’re interested in cash flow–because it stands to reason that cash (either now or in the future) is what any investor wants from an asset, whether it’s a stock or a piece of real estate. Cash flow for stocks is tough to pin down, but certainly the lower bound is dividends and the upper bound is roughly after-tax reported earnings.
Let’s focus on dividends. Much has been said in recent years about how low dividend yields are. That’s correct, but what’s little understood is that dividends have been growing–not just with inflation, but in real terms.
In fact, they have been growing in a spectacular fashion. Hassett and I did calculations using aggregate data and found that the average real growth rate of dividends since 1959 has been 5 percent! The assumption we used to derive our figure of a price-to-earnings ratio of 100 was a real growth rate of 2.1 percent and a nominal growth rate of 4.9 percent.
Now let’s focus on reported earnings. Consider General Electric Co. It currently trades at a P/E of 32, which is roughly an earnings yield of 3 percent. In other words, when you buy $1,000 worth of GE stock, you are buying profits today of $30 (roughly half of which will be distributed to you in the form of dividends, with the rest invested in the company, boosting the stock price over time).
Now, if GE increased its earnings merely at the rate of inflation, then 10 years from now, you would still be the beneficiary of only $30 in real profits–and you and Siegel would be right. Clearly, a stock that throws off 3 percent in real cash flow annually is inferior to a bond that throws off 4 percent.
But, instead, over the past five years, GE has been increasing its profits (and its dividends) at an average of 14 percent annually, according to my Bloomberg. Let’s assume that growth rate continues for the next five years. By 2003, your share of GE’s profits, on that original $1,000, will be $58–or a return, in that year, of 5.8 percent (the equivalent of a P/E of 17). In real terms (assuming 3 percent inflation), your share of the profits will be $50, for a real return in 2003 of 5 percent. Go out another five years and your nominal return in 2008 will be 11.1 percent, while your real return will be 8.5 percent.
In other words, by the 10th year you’ve held GE stock, your returns are twice the returns of Treasury bonds. By the 20th year, the real return on GE stock will be 24 percent–or six times the real return on Treasury bonds, at a level of risk that is essentially the same (according to Siegel’s research).
Is GE an anomaly? Maybe. So just assume that real returns are what we project them to be: 2.1 percent. Do the math and you find that, with Treasury rates where they are today (5.9 percent for 30 year bonds), a P/E of 100 is fully justified.
Now, I don’t suggest that you necessarily run out and buy stocks that are trading at a P/E of 32. Why not buy Caterpillar Inc., at a P/E of 13? Estimated annual earnings growth for the company over the next five years, according to Zacks Investment Research, is 10.3 percent. Thus, Caterpillar, which is currently producing an earnings yield of 6.3 percent, should have a yield of 10.1 percent in five years.
This is our point. If the traditional, irrational risk premium for stocks disappears, then stocks need to yield only 1 percent today (a P/E of 100) to produce the same cash flow, over a long period, as T bonds yielding about 6 percent. We are more than happy to buy stocks that yield 4 percent (a P/E of 25).
If you and Siegel are right about earnings rising merely with inflation, then obviously our calculations are wrong. But earnings rise much faster. Why? That’s something we could discuss, but it’s undeniable that a) real growth has been the hallmark of modern capitalism and b) stocks, over the past 70 yeas, have returned a stunning 11 percent annually on average, which is far too much for their actual risk. Something has to give, and we believe the most likely course for the market will be up and up, until it reaches a level where annual returns will become realistic–i.e., much lower than they are today.
Anyway, maybe we’ve talked enough about me. How about you? How are you smart enough to know that the United States is suffering a “bubble economy”?
Sincerely,
Jim
