Created 12/7/1996
This is
Brad De Long's Home Page

Is the Stock Market Overvalued?

By J. Bradford De Long

This is a rough draft. The published, edited, and (much) improved version is in Slate (December 21, 1996).

Greenspan Shakes the Market

Federal Reserve Chair Alan Greenspan briefly shook world financial markets the night of Thursday, December 5, with a single question: "How do we know when irrational exuberance has unduly escalated asset values?"

From Asia to Europe to New York, traders interpreted Greenspan's delicate rhetorical query in an evening speech at a right-wing think tank to imply that the stock market was overvalued, and that the Federal Reserve might be about to increase interest rates to dampen the market. So investors bailed out over the western hemisphere's Thursday night, pushing stock prices down between 2 to 4 percent. In the aftermath, Federal Reserve officials hastened to assure the press that a rise in interest rates was not in the offing, and that Greenspan's expressions of concern were merely ... expressions of concern.

Is the stock market overvalued? Would a stock market crash put the economy in peril? What was Alan Greenspan trying to do that Thursday evening? And--if the market truly is overvalued--should the Federal Reserve be trying to talk stock prices down by expressing concern, or to push stock prices down by increasing interest rates?

The stakes for investors are truly enormous. If the $7 trillion U.S. stock market is overvalued by a third, some $2 trillion plus of the wealth Americans now hold in stocks will vanish over the next decade as stock prices return to fundamentals. The losers will be those who remain fully invested in the market over the next decade. If stocks are not overvalued today, the losers will be those who--out of fear of possible overvaluation--spend the next decade out of the stock market, with their wealth invested in lower-return investments in bonds and in the money market.

Irrational Exuberance?

No one disputes that stock prices are very high.

One standard measure of "fundamentals" is average earnings over the past 10 years--an average taken over a time period long enough to smooth out business cycle fluctuations in profitability. In a typical year, a typical stock is priced at about fifteen times its 10-year average of earnings. Today the typical stock sells for nearly thirty times its 10-year average of earnings.

The argument that the stock market is overvalued--and that it will come back to earth over the next decade, perhaps in a gradual deflation of prices like a slowly-leaking balloon and perhaps in a crash--is simple. Stocks are tradable pieces of paper that carry "ownership" of the earnings of American corporations. Stock price "fundamentals" are thus roughly proportional to the earnings of American corporations. But today the stock market is selling for roughly twice its typical earnings multiple.

What goes up comes down: sometimes rapidly, sometimes slowly. In the past, whenever stock prices have gotten as high relative to fundamentals as they are today the subsequent decade has been an extremely bad one in which to invest in the stock market. Years in which the ratio of prices to ten-year average earnings has been above twenty have seen subsequent stock returns over the next decade lower than investments in bonds promise today. There have been no years in which the ratio of prices to ten-year average earnings has been as high as it is today.

On the Other Hand...

On the other side of the issue, bull marketeers offer three main theories for why current stock prices are not too high:

First, some claim that the information technology boom is ushering in a generation-long boom of rapid growth and rising profits. Past rule-of-thumb valuations based on earnings and dividends assume that economic and profit growth will continue in the future at roughly the same pace at which it has proceeded in the past. But because we are on the threshold of the post-industrial transformation, there is no good reason to think that ecoomic growth--and earnings growth, and dividend growth, and stock price growth--will be faster in the future than in the past.

Second, others claim that the rate-of-return that the average investor expects to receive on stocks has fallen. In the past demand for stocks was limited by fear of risk. Investors could look over at the bond market, see the returns available to them if they invested in bonds or in the money market, and think: "Investing in the stock market seems to be much more risky than investing in the bond market; I'm not going to run that risk unless I think stocks are very attractively priced and offer relatively high dividend yields." Thus demand for stocks was relatively low.

But, these analysts claim, the inflation of the 1970s inflation taught investors the brutal lesson that there is no safety in investments in bonds: your investment in the "safe" bond market can vanish if interest rates rise, or if inflation devalues the purchasing power of your bond portfolio's principal. Investors today know that there is no safety in bonds--and have less fear of the business-cycle and other risks associated with investments of stocks. Hence demand for stocks today is higher than in past generations, and higher relative demand means higher stock prices.

Third, over the past generation corporations have learned better how to avoid paying taxes. Corporative executives today manager their firms' capital structures in ways that amount of do-it-yourself expropriation of what used to be the government's share of corporate profits. Some companies have pushed up their debt-equity ratios, and thus changed payments to investors that used to be called "dividends" into "debt interest." Dividends are paid out of earnings, and before they are paid the government takes its cut through the corporate income tax. Debt interest is a cost, paid out of pre-tax operating cash flow. The net effect is to shift some of the cash flow from the enterprise away from the government and to private investors.

Other companies have decided to reduce their cash dividends and use money that would have been earmarked for dividends to buy back shares instead. Suppose you buy back some of each investor's shares and then split your stock proportionately; afterwards your investors have the same number of shares of stock with which they started, and they have the cash committed to the buyback as well. Suppose you take the buyback money and pay it out as a cash dividend to your investors; afterwards your investors have the same number of shares of stock with which they started, and they have the cash that would have been committed to the buyback as well. Is there a difference? Yes. Income earned as dividends is taxed as ordinary income, while income earned through a buyback is taxed at the lower capital gains rate. More important, perhaps, is that investors choose when and if to sell their stocks, and thus when and if to realize capital gains and pay capital gains taxes taxes. This option to defer taxation until a favorable moment is extremely valuable to taxpayers. Thus the stock of companies that buy back shares should be more valuable than the stock of companies that take the same cash and pay it out as dividends.

Are any of these three theories correct? To some degree yes: we can all find cases and situations in which high valuations today are the result of success at do-it-yourself expropriation of the government's share of the enterprise, or of high expected growth due to anticipation of the post-industrial revolution. All recognize--after the experience of the 1970s--that investments in bonds are extremely vulnerable to the risk of even moderate inflation.

Are their effects large enough to justify the current high price of stocks? Perhaps. There is no certainty in finance. J.P. Morgan, the turn-of-the-century financier whose name still echoes through American finance in J.P. Morgan and Morgan Stanley, had a stock answer when asked him what the stock market would do: "It will fluctuate." Lloyd Bentsen has a stock answer when asked what financial markets will do: "If I knew, I wouldn't be standing here. I'd be calling from my yacht."

At the peak of every previous bull market, there has been a vocal and vociferous group of market authorities and prognosticators assuring investors and the public that the old rules of thumb are no longer valid. The most famous of these predictions came just before the stock market crash of 1929, when Yale professor Irving Fisher reassured investors that stock prices had attained a "permanently high plateau."

Every time up until now, the vocal and vociferous authorities claiming that dividend growth will be faster or required rates of return lower than in the past have been wrong. And what had gone up has come down--most notably in the early 1930s and the early 1970s.

Maybe this time will be different.

What Is To Be Done?

What if the stock market is too high? What if it is undergoing an episode of what Alan Greenspan would call "irrational exuberance"?

Irrational exuberance in financial markets leads to what investors have for nearly three centuries called "bubbles"--overvalued markets--because they can vanish suddenly when pricked. Bubbles do not always burst in a sudden crash, but they often do.

Financial havoc may follow a crash as it disrupts the web of confidence and promises that keep financial markets running smoothly. Banks refuse to loan to businesses they fear are insolvent. Depositors refuse to deposit their money in banks they fear are bankrupt. The chain of financial transactions that transforms incomes into savings into business purchases of plant and equipment that employ workers collapses.

After the financial crash of 1929 and the large-scale bank failure of 1930-1933, the U.S. unemployment rate rose to 25 percent, nearly five times its current level. Following the late-1980s crash of the Japanese stock and property markets, the Japanese economy stagnated. Japanese incomes and production would be one-quarter higher than they are today if that country's 1990s economic growth matched its 1980s pace.

But prolonged depression is not inevitable after a financial crash. When stock prices fell by a quarter in one day in 1987, the American economy barely noticed. The 1987 market crash had little effect on the economy in large part because Alan Greenspan and company headed off the destructive chain of bankruptcies-and-defaults before they could start. Even so, it is better to avoid the crash-triggering bubble than it is to try to keep a crash from triggering a depression.

If we grant for the moment that the stock market is overvalued, what then should the Fed do?

One answer is that the Federal Reserve should do nothing. The history of central bank attempts to deflate overvalued stock prices is not encouraging. When the Fed tried to try to cool off stock prices in 1929 it had no impact on the stock market, but it did set in motion the depression it had hoped to avoid. The Fed's principal stock-deflating tool is an increase in interest rates, which draws money out of the stock market and into the bond market.

But raising interest rates now (which would depress the economy now) to avoid a possible financial crisis (which would depress the economy later) is a lot like destroying the village in order to save it.

Maybe a second answer is that the Fed should express concern, as Greenspan did. Such expressions might subtly shift market psychology and begin the gradual deflation of the bubble. The risk is that the shift in market psychology might not be subtle, and the deflation of the bubble might not be gradual. This too would bring on the crisis that the Fed would wish to avoid.

What Did Greenspan Do?

Note how gingerly Greenspan expressed his "concern." He did not say that he was worried because the stock market was overvalued; he asked how he could figure out whether the stock market was overvalued. He did not say that he would take any action in response to overvaluation; he asked if monetary policy should be any different if there were overvaluation.

He did express concern. And Federal Reserve officials working from behind the formal Washington screen of official anonymity did confirm that Greenspan had meant to express concern, and had been more than musing out loud. But as John Berry of the Washington Post reported, Greenspan's wording was so tentative that "even some other Fed officials, who had seen drafts [of the paragraph] over the past month, did not realize it was intended to be a signal that the chairman thought that stock prices had become uncomfortably high."

It is hard to imagine a smaller step than the one that Alan Greenspan took: a small rhetorical question, wrapped in a paragraph confessing uncertainty, inside a speech of philosophical musings about the art of cental banking, given in the evening, made before not a financial audience but a right-wing think tank.

John Kenneth Galbraith's history of the crash of 1929 reports that it was clear that market psychology was unstable, and a crash a clear possibility, after the so-called "Babson break"--a sharp but temporary decline caused by nothing more than a casual prediction that the market might decline. Earlier this month we had a Greenspan break--a sharp but temporary decline caused by nothing more than a rhetorical question about how we would recognize a market dominated by "irrational exuberance."

J. Bradford De Long is associate professor of economics at the University of California at Berkeley and co-editor of the Journal of Economic Perspectives, the intelligible journal of the American Economic Association.

From 1993 to 1995, he was deputy assistant secretary of the Treasury for economic policy.


Created 12/7/1996
This is
Brad De Long's Home Page

Professor of Economics J. Bradford De Long, 601 Evans
University of California at Berkeley
Berkeley, CA 94720-3880
(510) 643-4027 phone (510) 642-6615 fax