Dow 36,000: Future or Fiction?Published: October 27, 1999 in Knowledge@Wharton
The sustained boom on Wall Street has recently resulted in a spate of books with such titles as Dow 36,000, Dow 40,000 and even Dow 100,000. Coupled with euphoria over the coming of the Internet age, such monikers sometimes suggest that the stock market will keep soaring into the stratosphere well into the next millennium. But how realistic is it to believe that the Dow Jones Industrial Average, arguably the best-known yardstick of Wall Street’s performance, will keep defying gravity?
Jeremy Siegel, professor of finance at Wharton and author of the well-known book Stocks for the Long Run, recently discussed these issues with Kevin A. Hassett, an economist at the American Enterprise Institute and co-author with James Glassman of Dow 36,000: The New Strategy of Profiting from the Coming Rise in the Stock Market. Hassett began the discussion by summarizing why he and Glassman think the Dow, which now hovers around 10,000, could rise to 36,000.
Hassett: Assume you want to float a junk bond. In order to convince people to lend you money, you must pay them a high interest rate. A 30-year Treasury bond may have a yield of 6%, but for your high-risk bond you may have to pay 12% to convince investors to give you money. The difference, 6%, is called the risk premium.
Now suppose that suddenly, everybody decides that your hypothetical junk bond isn’t really as risky as people thought--and that it is really a blue-chip company. Then those who had bought the bond before the news came out would get a big capital gain, and the yield on the bond would go down to, say, 7%.
In Dow 36,000 Glassman and I argue that a similar phenomenon has been going on in the stock market. People have realized that over the long term, stocks aren’t as risky as they thought, and this has forced down the equity risk premium. Jeremy Siegel has a paper where he shows better than we do that this process has occurred. The key question is: Is this process over, and have we reached the point where the equity risk premium is as low as it gets? Or is there room for it to go down further?
This is a controversial question. Our calculations suggest that the current equity risk premium is around 3%. Others using different techniques get answers in the same range. If the equity premium is in the 3% to 4% range, Jim and I think that there is room for it to go down further. We don’t mean that it has to go down, but we think that it could—for two reasons. First, for long-term investors, stocks aren’t nearly as risky as bonds. So why should stocks pay a premium? We think that they shouldn’t—their equity premium should be close to zero. Secondly, for short-run investors, a lot of evidence has convinced me that the reasonable equity premium is between half a percentage point and 1%.
Glassman and I think that if people are perfectly rational—and the equity premium for long-run investors is around zero, and for short-run investors it is between one-half and 1%, yet they are buying stocks at an equity premium of 3% to 4% -there is plenty of room for good news. The market has plenty of upside.
Siegel: Of all the bullish books about the stock market, the one by Glassman and Hassett is by far the best grounded. They use good theory, they use findings both from academia and from the investment profession, and they are not trying to invoke any New Age, New Era thoughts about the world moving towards utopia. Their work deserves to be taken seriously. My conclusions do differ from theirs, but my differences with them are not based on the fact that they misquote or misinterpret my studies. They are meticulously accurate in taking my findings. But I have some problems with their analysis.
The basic concept they use from my book, Stocks for the Long Run, is that I found that over long periods, of between 15- and 20-year periods or more, stocks are no riskier than or might be safer than standard bonds once the inflation risk is taken into account. The key here is the inflation risk: It’s not just the number of dollars that these instruments provide, but their purchasing power over long periods of time. Glassman and Hassett say that if that is true, then stocks should not be priced any differently than bonds.
This raises some interesting issues. I would be flattered if the world thought my results are as convincing as Glassman and Hassett do, and the market priced stocks and bonds as being equally risky. But even some of my colleagues don’t agree with me. They say that while they do not question my data, I do not have enough independent long-term periods to make a statistically significant statement.
Hassett: I agree. You should be cautious in using evidence like that. Our book is full of examples that highlight why the risk for stocks is declining. For instance, Chrysler was selling for $2.81 in 1977 and paying a dividend of 20 cents a share. Right before the Daimler-Chrysler merger, Chrysler was paying $1.60 a share. A little bit of dividend growth every year made it so that relative to Chrysler’s paid-in capital, its yield 20 years later was 57%. In the aggregate, firms have not grown quite as remarkably, but they have grown quite a bit. Glassman and I think that the cash flow from firms grows as long as the economy is growing. Einstein’s most powerful force in the universe--compounding--provides the impetus for the stock market’s growth. The market can fluctuate, but the yields can not be as high as 57%. That is why we believe that stocks’ risk in the long run is softened considerably. It has happened because economic growth and the growth of profits has been steady and because compounding is such a powerful force.
Siegel: I agree. Stocks are claims on real assets and so they have held up through crises much better than monetary assets. This is particularly true internationally. So I can appeal to theory to tell you that stocks in the long run would be as safe if not safer than bonds. A lot of theory suggests this result.
Let me mention another angle—whether the risk of the long-term holder will ultimately be priced into the market. Suppose the world accepts that the long-term risk of stocks is no greater than that of bonds. That applies to the long-term holder. What you are saying is that the pricing of these securities is now going to be done by the long-term holder. The question is: Is there enough long-run money in the market to price it to that level?
Hassett: We dealt with that issue in the first Wall Street Journal piece that we wrote about our findings. Let me read you a quote. After writing about Dow 36,000, we said: "That may overstate the case. For the premium to vanish, investors must hold their shares in a diversified portfolio for at least 20 years. In practice, investors often waver."
Siegel: That’s accurate. But to get to your numbers, these long-term holders must be the ultimate pricers in the market. Let me now move to a related question. This is another difference between the Glassman-Hassett hypothesis and mine. The way you get your result is that you look at the pricing of long-term bonds and say that the stock market will move to the returns on bonds. While I agree that stocks have been underpriced through history, I also think that bonds have been overpriced. Instead of the real return on stocks going down to that of bonds, they will both come together at some real return that is in between the two. The real return on bonds is likely to rise and the real return on stocks will go down. I don’t think the gap will be eliminated, because I don’t think that long-term investors will control that market. But the gap will be closer. I think the real return on bonds will go up from the 1%-2% range to about 4%, and the real return on stocks will go down from 7% to about 5%. Given that, my feeling is that the market has already made a large move to that level. We see the price-earnings (P/E) ratio on the S&P 500 has risen to 30. Looking at the earnings yield on the market, I think lower yields are already being priced in to some extent. The current P/E ratio is 30, and since the historical average has been 14 or 15, we are twice as high as the average of the past 130 years.
Hassett: Your point that the bond rate should move up a little—so that you don’t go to Dow 36,000 but to Dow 22,000 or something like that—is reasonable. No one understands real interest rates well enough to rule that out. It is true that except for the last few months, there has been no discernible long-term trend in real interest rates. But there is a discernible long-term trend in the stock market. The stock market has been going up, while the interest rate has been fluctuating around in the 3% range as long as we have measured it.
Siegel: Let me now turn to another crucial question. In January 1997, for the first time in U.S. history, the Treasury issued inflation indexed bonds called Treasury Inflation-Protected Securities or TIPS, where all coupon payments and final principal are indexed to the consumer price index. This means the yield that you receive on that bond is a real yield. In other words, there is no inflation risk on these bonds; the government automatically compensates you for changes in the price level. Many other countries have such bonds; the U.K. has had them for 20 years and Canada has had them for 10 years.
When these TIPS bonds were first floated, their real yield was just over 3%. That yield has been creeping up slowly but surely and has now reached 4.13%. That is astoundingly high.
Hassett: We tell everyone in the book to buy these.
Siegel: I tell everyone to buy them too. But I still say that their yield is not as good as stocks, and investors will still do better in stocks over the long run. If you take these bonds as the safest risk-free asset, stocks have to be priced up from there. Stocks have to have a higher yield. With the 30 P/E ratio we have now, it is not going to be easy for stocks to push above that 4% unless we have better than long-run earnings growth or other special factors.
Hassett: This is a really important point. TIPS are very thinly traded right now, and if everybody did take our advice and buy these bonds, that would drive the yield down significantly. These bonds are as unpopular as they are ever going to be.
Siegel: There is some truth to that. Most economists think that TIPS are underpriced, so that yield of 4.1% is too high. But is it not true that if I were to use the current TIPS yield, your analysis would not find any increase from the current level of the Dow?
Hassett: No, you would get some increase.
Siegel: Are you sure? The Dow Jones yield is 1.5% and you use a 2.3% real dividend growth.
Hassett: We did the calculations in lots of different ways in the book. The problem with calculating the value of the market based on how much money firms generate is that that old-fashioned measure, the dividend, is getting worse and worse as a measure of firms’ earnings. The dividend payout rate used to be 70% of earnings in the 1950s and now it is just 30% of earnings, mostly because firms are getting better at managing taxes. So you have to adjust the dividend yield somehow when you think about how much a firm is worth, and the question is how. We do it a zillion ways.
Siegel: What the readers can see is that your results depend critically upon such numbers.
Hassett: It is explored very responsibly in the book. One of my favorite sets of calculations is where we taper off the recent growth at different rates. We assume that we get the present rate of growth for the next five years and then it tapers off to 2.5%. After that, we assume high growth for 10 years and 20 years before it tapers off. It is the collective inspection of all those numbers that adds up to 36,000, if bonds don’t move.
Siegel: Let me shift the argument a little. A big part of the book talks about PRPs or perfectly reasonable prices, which is an acronym for the right price or a rational price for these securities, given your assumptions. You also say that we should not be afraid of pricing the market—not an individual stock, but the market—at 100 times earnings. This is quite dramatic. The P/E ratio during the past 130 years has averaged 14. That may be too low, and the market was underpriced, but 100 is quite extraordinary. Until the Internet stocks came around, there had never been big cap growth stocks that achieved 100 P/E ratios. I have trouble with people saying they will pay 100 times yearly earnings of a stock. It sounds to me like everyone should quit their jobs and form corporations and float shares on the market.
Hassett: If you can start the next Cisco Systems, you should definitely do it right now. But if you can’t do it, you should buy their stock.
Siegel: Cisco right now is selling for about 80 times earnings, and it is relatively high. But you say this about the whole market, not just about the Microsofts and the Ciscos.
Hassett: Cisco is a striking example, and striking examples can be very instructive. Cisco at the time we completed our book was selling at $64 a share and it had earnings per share of 74 cents. The P/E ratio was about 85. If Cisco grows its earnings for the next five years at the same rate it has for the past five—and the company shows no signs of trailing off—it will increase its earnings by a factor of 10. If Cisco’s price doubles over that time, its P/E ratio will be 17.
Siegel: There is nothing wrong with what you are saying. But the key question is whether all the capital in the market—not just Cisco—will have a P/E ratio of 100. The growth rate of other corporations’ earnings is going to be nothing like Cisco’s.
Hassett: That’s right, but that is where the valuation comes from: The highly valuable U.S. firm. Don’t forget that when you do these averages, you are doing a market-cap weighted average, and the winners get much more weight than the losers. The average market cap weighed firm in the U.S. has a significant ability to borrow a dollar at 6% or 7% and earn 10% or 15% on that dollar. That gap between what they make and what it costs them to get the money is why earnings and dividends per share go up over time.
Siegel: There may be a group at the top whose P/E ratios are 200 or 300, but there will also be the rest, which will be down at, say, 20.
Hassett: You are right. Every firm isn’t at the level where it should be traded at 100. We list five questions investors should ask before they buy a stock to help our readers figure out which will be the winners. The first question is: Does the firm make money?
Siegel: Does that mean we should scratch most of the Internet firms? Right now only a couple of them are making money.
Hassett: I don’t want to own firms that can’t answer yes to these questions. I am a conservative investor. Making money is tricky. Benjamin Graham in 1949 warned speculators that while airplanes had begun to fly everywhere, he did not see how airline stocks were ever going to make money. And he was right. Question 2 is: Does the firm have something special? We give examples that are very well known, such as Coca-Cola or Tootsie Roll. These firms have something enduring that cannot be reproduced.
Siegel: That almost sounds like a Warren Buffett criterion.
Hassett: The next question is: Does the firm have a long-term history of earnings growth higher than 7.2%?
Siegel: Why do you take 7.2%?
Hassett: The reason is that if you plot a histogram, it’s right near the average, and the firm must have above-average growth. If a firm grows more than 7.2% a year, then it doubles in size in 10 years and we like that because it gives readers a useful rule of thumb. The next question is: Will the firm exist 50 years from now? You may never know that answer for sure.
Siegel: You could also ask if the firm that swallows it up will be around for the next 50 years. The only firm from the Dow 30 that is left from the original group is General Electric.
Hassett: There is an actual example of a mutual fund called Lexington Corporate Leaders that organized itself around this question. It was founded in 1935 during the Great Depression. It charged an upfront 12% load and the intention was to pick 20 stocks that would last until 2015 when the fund would end. The fund has firms like GE, Sears, AT&T, Procter & Gamble, Eastman Kodak, and others that will probably last for another 100 years. That is why it is important to ask how long these firms will be around. If they are, compounding will ensure that those companies will make a lot of money in present value.
Siegel: You are basically using a growth criterion. Do you think that these growth stocks should be priced in the 200 P/E range or so?
Siegel: In my book I analyzed the big cap growth stocks of the early 1970s, and I was surprised how well they had done. One thing that did strike me, however, was that the highest P/E stocks among them generally did not keep up with the market. Polaroid was the highest, with 95, and it has done miserably since. I don’t want to go down the list, but none of the stocks that actually had a P/E of more than 50 were able to match the market, though some that were just below 50 did much better than the market.
One thing that is very important for people to know is that in your brave world of Dow 36,000 the forward-looking returns of stocks are going to be miserable. Stock prices are going to be bid up to the level they should be, and after that they won’t do much better than the old bonds. That is supposedly going to be our fate for the rest of the millennium. I have a hard time believing that people will put their capital at risk for 2.9% or 3% real returns, which is what is supposed to happen in the world of Dow 36,000.
Hassett: There will still be a lot of gains for things like tax deferral and so on. We have used very conservative and low growth assumptions. The key point is that the force driving the market up through today has mostly been the decline in the equity premium—like the junk bond I mentioned in the beginning. Going forward, there is no reason to believe that the decline will reverse itself, which is what the so-called bubble-ologists are saying. There might even be reasonable arguments for it going much further down. If the equity premium goes from 3 to zero, then the market has to go up. Exactly how far up it will go is tricky to assess because little changes in numbers make a big difference.
Siegel: Let me say in summary that I do believe that the equity premium has come down from what it has been historically. I also believe that in the long run, the risk of stocks is as low if not lower than nominal bonds. But I think that the market has already made a very substantial correction towards those facts. We are closer to the end of the adjustment rather than the beginning. You might probably say that we are at the beginning.
Hassett: Actually we are in the middle. The Dow has gone up by a factor of 13 since 1982. We say in the book that it has to go up by a factor of three to finish the job. If you look at it in that perspective, that’s all you need to get to Dow 36,000.
Siegel: That depends on how far you go back. When the Dow was at 780 in 1982, that was its second-most undervalued position in U.S. history, outside of the panic of a few weeks during the Great Crash. But during the 1980s a lot of good things happened. Inflation was down, the monetary policy was under control and growth was up. A lot of these factors have continued into the 1990s. As grossly as the market was undervalued in 1982, it came up to historical values in the early 1990s. In the late 1990s it has gone to a level that is discounting many good things as well as a reduction in the premium. The question now is how much further the adjustment has to go. I think it is now near the end.