NBER WORKING PAPER SERIES

A SHORT NOTE ON THE SIZE OF THE DOT-COM BUBBLE

J. Bradford DeLong
Konstantin Magin

Working Paper 12011

[Link]

NATIONAL BUREAU OF ECONOMIC RESEARCH

1050 Massachusetts Avenue
Cambridge, MA 02138
January 2006

The views expressed herein are those of the author(s) and do not necessarily reflect the views of the National
Bureau of Economic Research.

©2006 by J. Bradford DeLong and Konstantin Magin. All rights reserved. Short sections of text, not to
exceed two paragraphs, may be quoted without explicit permission provided that full credit, including ©
notice, is given to the source.
A Short Note on the Size of the Dot-Com Bubble
J. Bradford DeLong and Konstantin Magin
NBER Working Paper No. 12011
January 2006
JEL No. G1

ABSTRACT

A surprisingly large amount of commentary today marks the beginning of the dot-com bubble of

the late 1990s from either the Netscape Communications initial public offering of 1995 or Alan

Greenspan's "irrational exuberance" speech of 1996. We believe that this is wrong: we see little

sign that the aggregate U.S. stock market was in any way in a significant bubble until 1998 or so.

J. Bradford DeLong
Department of Economics
Evans Hall, #3880
University of California, Berkeley
Berkeley, CA 94720
and NBER
jbdelong@[Link]

Konstantin Magin
COINS
McLaughlin Hall, #1726
University of California, Berkeley
Berkeley, CA 94720
magin@[Link]
 A Short Note on the Size of the
Dot-Com Bubble
J. Bradford DeLong
U.C. Berkeley Economics and NBER

Konstantin Magin1
U.C. Berkeley COINS

January 2006

Abstract: A surprisingly large amount of commentary today

marks the beginning of the dot-com bubble from the late 1990s as
the Netscape Communications initial public offering of 1995 or
from Alan Greenspan’s “irrational exuberance” speech of 1996.
This is wrong: there is little sign that the aggregate U.S. stock
market was in any way in a significant bubble until 1998 or so.

Introduction
Most finance economists will find this short note to be obvious and trivial. But

many others will find it surprising and useful. Our message is simple: from our

perspective, perhaps the most interesting thing about the late-1990s dot-com stock

market “bubble” is how short it turned out to be. And yet many today—much of

the conventional wisdom we hear in the financial, business, and political press—

takes the bubble of the 1990s to have been much larger and longer than it in fact

was.

1
Part of this research was supported by the National Science Foundation through
the U.C. Berkeley COINS center.
The Netscape Initial Public Offering

Part of the conventional wisdom dates the start of the bubble to the extraordinary

excitement occasioned by the Netscape Communications, Inc., initial public

offering in March 1995. Advertisements for the PBS television show Frontline’s

report on the stock market and the internet—“Dot Con”—say that the show:

… traces the rise of the bubble from the day in August 1995 when
a young company called Netscape Communications went public….
Within a few hours of the market's opening on Aug. 9, 1995,
Netscape's stock shot from $28 to $75 per share, closing at $58.25
at the end of its first day of trading. It was an historic—and
prophetic—moment on Wall Street. "Nobody expects what
happens at Netscape," says Joe Nocera, executive editor at Fortune
magazine…. “And suddenly, if you're an investment bank, you
realize that this is something that can be taken advantage of." (see
Frontline (2002)).

Business commentator Maynard Patton (2005) calls the day of Netscape’s initial

public offering, “Day One of the Internet Bubble.” The New Yorker’s John

Cassidy (2002) refers to the initial IPO of the South Sea Company in 1720—the

first time “bubble” was used in English in its current meaning—as “Netscape-

like.” San Jose Mercury News internet columnist Dan Gillmor (2005) calls the

Netscape IPO “the Big Bang of the Internet stock bubble.” CNN Insight’s Bruce

Francis says that it was the “unusual high demand for Netscape's offering [that]

was… the birth of a bubble.” And Forbes’s Quentin Hardy writes that it was the

“Netscape IPO… [that] kicked off dot-com mania on Wall Street.”

2
But as of the moment of the Netscape initial public offering, the stock market was

not then in a “bubble.” Even the tech-heavy NASDAQ index was not then in a

bubble. Those who invested in the NASDAQ at the moment of the Netscape IPO,

in August of 1995, and thereafter shadowed the index have earned real returns

averaging 9.3% per year from then to November 2005. Those who invested in

Netscape Communications at its closing price on its IPO date of August 9, 1995

and who held on to their stock in Netscape until the company’s absorption by

AOL in 1999 earned an average real rate of return of 35% per year on their

investment.

Greenspan’s “Irrational Exuberance” Speech

A second part of much conventional commentary dates the start of the dot-com

bubble to sometime before Federal Reserve Chair Alan Greenspan’s December 5,

1996 “irrational exuberance” speech. Speaking before an after-dinner audience at

the Washington D.C.-based American Enterprise Institute, Greenspan posed a

rhetorical question: "How do we know," he asked, "when irrational exuberance

has unduly escalated asset values?" Greenspan’s speech led the stock markets in

Tokyo and Hong Kong to fall by 3%, the markets in Frankfurt and London to fall

by 4% when they opened, and the U.S market to open down 2%. Greenspan’s

words had powerful effects, either because his view of fundamental values was

3
respected or because investors feared his rhetorical question was the prelude to a

tightening of monetary policy.

Indeed, the odds are that Greenspan did believe the market suffered from

irrational exuberance when he gave his December 1996 speech. As Wall Street

Journal reporter Jacob Schlesinger (2000) reported, two days earlier:

… having watched the Dow surge a dizzying 27% that year,

Federal Reserve Board Chairman Alan Greenspan hosted a private
meeting…. On one side was Abby Joseph Cohen… who came
from her post at Goldman, Sachs & Co. to defend investor sanity.
She methodically gave Fed governors a list of reasons why
underlying economic changes justified such lofty prices in the
market. On the other side were two Ivy League economists, Yale's
Robert Shiller and Harvard's John Campbell, who painted a much
gloomier picture, though they didn't address Ms. Cohen's
comments directly. They illustrated their message of portent in 10
pages of handouts showing trends going back to 1872. The markets
were destined, at best, to tread water, and possibly to crash, they
warned…

Since 1994, Schlesinger (2000) reported, Federal Reserve staff economists were

forecasting that a stock market correction was likely: “Some of the economists

told colleagues that they had personally gotten out of the market, and advised

others to follow them into bonds. Mr. Greenspan privately shared many of these

doubts…. In February 1994, for instance…. ‘We partially broke the back of an

emerging speculation in equities’, Mr. Greenspan contentedly told his colleagues

in a conference call…. ‘We had a desirable effect’."

Federal Reserve Chair Ben Bernanke, at the Federal Reserve’s Jackson

4
Hole Symposium last August, said that it was primarily “the stock market

bubble which led to an unsustainably high level of economic activity and tax

revenues" that was principally responsible for the improvement in America’s

fiscal position in the 1990s (see Henderson (2005)). In mid-1996, Robert Shiller

(1996) wrote about the long-run stock market outlook, saying that “it is hard to

come away without a feeling that the market is quite likely to decline substantially

in value over the succeeding ten years; it appears that long run investors should

stay out of the market for the next decade,” although Shiller also warned that

“the conclusion of this paper that the stock market is expected to decline over the

next ten years and to earn a total return of just about nothing has to be interpreted

with great caution… [dangers of\ data mining… [possibility of] structural

changes… that mean that the past of the stock market is no longer a guide to the

future.

Yet as of the moment of Greenspan’s speech, the NASDAQ index stood at 42%

of its current value. If you had adopted a strategy of buying, holding, and

reinvesting in the NASDAQ index when Alan Greenspan made his “irrational

exuberance” speech, from that moment until the time of this writing at the start of

October 2005, you would have realized a real return of 8.1% per year. The answer

at the time to Greenspan’s question—was the stock market in late 1996 at values

that had been unduly escalated by irrational exuberance—was “No.”

5
 Figure 1: Nominal Value of the NASDAQ Index

Source: Datastream.

Table 1: Real Rates of Return for Investments in Largest High-Tech

Companies since December 1996


 
  
  
  
  
  
  
  
  
!  
"  
#$  

Source: Datastream.

6
Even a year later is probably too early to date the advent of the bubble. Investing

in December 1997 in the NASDAQ yields real rates of return to date of 6.4% per

year. From the standpoint of today’s valuations, the ten high-tech companies in

America with the largest valuations as of January 1998—Microsoft, Intel, IBM,

HP, Cisco, Motorola, Dell, Oracle, Texas Instruments, and Sun—appear to have

been substantial bargains. The companies with the largest market capitalizations

are those that are most likely to have had their values affected by “irrational

exuberance”: yet for six of the ten companies real returns since the beginning of

1998 have been more than ten percent per year, and for only two of the ten

companies have real rates of return been negative. American high-tech was, it

turns out, a good investment even as of early 1998.

There Was a Bubble

We do not want to dispute the claim that the U.S. stock market was in a serious

and significant bubble in some period leading up to its peak in the late winter of

2000: there is next to nobody will deny that the stock market underwent a bubble.

Let us briefly mention three facts that make us certain that the late-1990s stock

market was grossly overvalued.

First, the stock market was unable to figure out that 3Com was the majority owner

of Palm, and thus that 1.5 times the value of a share of Palm was a plausible floor

7
to any rational valuation of a share of 3Com—or alternatively that 2/3 the value of

a share of 3Com was a plausible ceiling to the value of a share of Palm (see

Cornell (2000); Lamont and Thaler (2001)). Parent company 3Com sold three

percent of its stake in Palm at Palm’s initial public offering, and promised that it

would spin off its remaining Pl shares by the end of 2000—with each share of

3Com owned entitling the shareholder to 1.5 shares of Palm. At the end of the

first day of trading, Palm’s closing price of $95.06 was not two-thirds but nearly

120 percent of 3Com’s closing price of $81.81. Lamont and Thaler (2001) report

that this is only one of six cases in the late 1990s in which the whole was less than

the sum of the parts: the parent firm was worth less than the subsidiary it had

promised to spin off. The Palm pricing anomaly lasted for more than two months.

Second, the NASDAQ index reached dizzying heights indeed as it exploded in

late 1999 and more than doubled in value in the year up to its late winter 2000

peak—without there being any plausible candidate for fundamental news to

support such a large revaluation of equity values.

Third, immediately afterwards came the huge and bloody bath taken by investors

in the NASDAQ from February 2000 to September 2002 as it lost three-quarters

of its value—again without substantial negative fundamental news. Note that the

NASDAQ underwent its bath even as the long-awaited recession proved

shallower than anyone had forecast, even as trend labor-productivity growth in the

economy as a whole proved to have accelerated faster than even the most rabid

8
new-economy boosters had dared to project, and even as the Federal Reserve

lowered the short-term nominal interest rates it controlled far and fast. The stock

market experienced a bubble.

When Did the Bubble Begin?

But how big was the bubble? And when did the late-1990s bubble begin? We

propose some extremely crude yardsticks:

Begin with the observation that the world has not been kind since 2000. In 2001

we had the terror-attack on the United States by Al Qaeda, which brought the

danger of widespread death and destruction materially closer. Less important, but

perhaps more relevant to equilibrium stock prices, is the fact that it appears to

have turned out to be more difficult to turn information and communications

technology technical excellence into durable profit flows than had been expected.

The major beneficiaries of high-tech innovation since 2000 have not been the

workers, entrepreneurs, and financiers of Silicon Valley, but of a different ilk:

think of Wal-Mart, its shareholders, and its customers (but not its workers) and

those like them as the real big winners from the high-tech new economy.

Taking account of these two pieces of fundamental news relevant to stock prices

that were unknown (and unknowable) in the late 1990s, we can set up our first

yardstick to assess the beginning of the bubble. Because the high-tech stocks in

9
the tech-heavy NASDAQ are risky, we would expect a rational market to price

the NASDAQ to produce returns higher than the 6.5% per year in real terms that

is the stock market's long-run historic average return. But we also suspect that

actual returns have fallen behind rationally-expected returns.

We guess that these two factors cancel out, and that the NASDAQ was in a

bubble--overvalued--when real returns from then until now have been less than

the 6.5% per year we expect from stocks. This is our first yardstick.

We also propose a second, more stringent, yardstick: define a "bubble" to be a

time in which subsequent returns do not match the three percent per year or so in

real returns than we expect on average from long-term bonds.

And we propose a third, most stringent, yardstick: define a "bubble" to be a time

in which subsequent real returns are negative.

When, according to these yardsticks, did the bubble begin?

According to both the second and third yardsticks—the more stringent ones—the

bubble was remarkably short. In October 1998 the realized NASDAQ real return

between then and now drops below 3% per year, and in November 1998 it drops

below zero. According to these yardsticks the NASDAQ was overvalued and in a

bubble for less than a year and a half before its peak in March 2000.

10
According to the first yardstick, the bubble was a year and a half longer. Since

April 1997, cumulative real returns on the NASDAQ have lagged the 6.5% per

year we expect from a diverified portfolio of stocks. Even using this yardstick, the

bubble lasted for less than three years before its peak.

Should This Come as a Surprise?

We do not believe that the short duration of the dot-com stock market bubble

should come as a great surprise. Macroeconomically significant bubbles are, we

believe, not common and not long lasting. John Kenneth Galbraith (1954) notes

that up until the second half of 1928 American stock market valuations were quite

respectable (assuming, of course, a normal future—one not containing a Great

Depression), and that only staring in late 1928 was speculation out of hand.

(DeLong and Shleifer (1991) put an even later date to the origins of the 1929

stock market mania.) Jeremy Siegel (2002) notes that the “Nifty Fifty” growth

stocks that led the bull market of the 1970s have turned out to be good

investments for sufficiently patient investors: “a portfolio of Nifty Fifty stocks

purchased at the peak would have nearly matched the S&P 500 over the next 26

years.” Indeed, the principal macroeconomically-significant puzzle for economic

theorists to be found in the stock market is not that stocks are occasionally grossly

overvalued and subject to bubbles because of irrational exuberance, but rather

than stock prices are so low as to generate extremely generous real stock returns.

11
It is the equity premium puzzle of Mehra and Prescott (2003) that is the most

significant macroeconomically-relevant feature of the stock market.

And we stress that just because the dot-com bubble was of relatively short

duration does not mean that it was of relatively small magnitude: the NASDAQ

index today is still only 40% or so of its peak value.

12
 References

John Cassidy (2002), [Link] : How America Lost Its Mind and Money in the
Internet Era (New York: Harper).

Bradford Cornell (2000), “The Parent Company Puzzle: When Is the Whole
Worth less than One of Its Parts?” (Los Angeles: UCLA xerox).

J. Bradford DeLong and Andrei Shleifer (1991), "The Stock Market Bubble of
1929: Evidence from Closed-End Funds," Journal of Economic History 52: 3
(September), pp. 675-700.

Bruce Francis (2000), “CNN Insight”

[Link]

Frontline (2002), “Dot Con”

[Link]

John Kenneth Galbraith (1954), The Great Crash (Boston: Beacon Press).

Dan Gillmor (2005), “The Netscape IPO a Decade Later”

[Link]

Nell Henderson (2005), “Rubin Praises Stance Of Greenspan on Deficits,”

Washington Post August 27, 2005
[Link]
4001:888164001&FMT=FT&FMTS=ABS:FT&fmac=&date=Aug+27%2C+2005
&author=Nell+Henderson&desc=Rubin+Praises+Stance+Of+Greenspan+on+Def
icits

Owen Lamont and Richard Thaler (2001), “Can the Market Add and Subtract:
Mispricing in Tech-Stock Carve-Outs” (Cambridge, MA: NBER Working Paper
8032).

Rajnish Mehra and Edward Prescott (2003), “The Equity Premium in Retrospect”
(Cambridge: NBER working paper).

Maynard Patton (2005), “Day One of the Internet Bubble”

[Link]

13
Jacob Schlesinger (2000), “How Alan Greenspan Finally Came to Terms with the
Market,” Wall Street Journal [Link]
May/[Link].

Robert Shiller (1996), “Price–Earnings Ratios as Forecasters of Returns: The

Stock Market Outlook in 1996”
[Link]

Jeremy Siegel (2002), Stocks for the Long Run 2nd ed. (New York: McGraw Hill:
007137048X%

14