Lesson 4 of 7
The dot-com bubble
Prices for companies without profits ran ahead of any cash they produced, and three losing years followed once the market asked for the denominator back.
By Javier Sanz
Founder of ValueScreener. Built Ninety Nine, a securities services provider, and sold it to a tier-one global bank; ran operations and internationalization at Alpaca, the brokerage API.
Who writes thisEditorial processLinkedIn (opens in a new tab)X (opens in a new tab)
In one minute
- Many of the era's most followed companies had no profit, so there was no denominator for a P/E, and prices were set on revenue growth or user counts instead.
- The S&P 500 with dividends fell in 2000, 2001 and 2002, three losing years in a row.
- A very high multiple is a claim about growth so large that even a right forecast can leave little room for error.
The dot-com bubble was a run of prices for companies that had not yet turned a profit, priced instead on how fast their revenue or their user count was growing. Once the market asked what cash those companies could produce, prices fell, and the S&P 500 fell with them for three years running.
Three losing years in a row
- 199921.0%
- 2000−9.1%
- 2001−11.9%
- 2002−22.1%
- 200328.7%
No profit, no denominator
A P/E divides a price by a profit. A company with no profit has no meaningful ratio to divide by, so the era's most followed names were priced on other things: how fast revenue was growing, how many users had signed up, how large the addressable market looked. None of those is cash the business has actually produced.
Three years, compounded
Falling 9.1 percent, then 11.9 percent, then 22.1 percent is not a 43.1 percent decline added up. Compounded, $1 becomes about $0.624 across the three years, a cumulative fall of roughly 38 percent. That is more than any single year on its own but less than the years simply added together. The S&P 500 holds large companies, many of them profitable but priced at very high multiples; most young dot-com names were never in it. A recession and the accounting scandals of 2001 and 2002 added to the fall.
What a very high multiple assumes
A share on 100 times earnings has an earnings yield of 1 percent: one dollar of profit for every hundred paid. If a buyer wants a 10 percent return, the other 9 points a year have to come from growth, indefinitely, for the price to be worth what was paid. That is not impossible for a young company for a few years. Asked to hold for decades, it is a demanding claim on a company that has not yet shown it can turn revenue into profit at all.
What was expected against what arrived
- Expected, Dec. 1999 survey19% a year
- S&P 500, 2000-2009, annualized−0.9% a year
Three warnings about the denominator
Shiller's cyclically adjusted P/E, built with John Campbell, averages ten years of earnings so a company does not look cheapest right at the peak of its own cycle. Marks's January 2000 memo, titled "bubble.com", asked whether the era's favoured companies could ever produce the cash their prices already assumed. Buffett's letter for 2000, written in February 2001, framed valuation around Aesop's three questions: how certain are the birds in the bush, when will they emerge and how many, and what is the risk-free rate. Shiller and Marks published before the fall; Buffett wrote after it. All three were asking about the denominator while the market was pricing the story instead.
Check yourself
Why did many of the era's most followed companies have no meaningful P/E?
- Their share prices were too low to compute one
- They had no profit, so there was no denominator to divide by
- Regulators had not yet required them to report earnings
Show the answer
Answer: B.
A P/E needs a profit in the denominator. A company that had not yet turned a profit had no meaningful ratio, whatever its price.The S&P 500 with dividends fell in 2000, 2001 and 2002. Compounded, roughly what was the cumulative fall?
- 43 percent, the three years added together
- About 38 percent, since a later fall compounds on an already smaller base
- 22 percent, counting only the largest single year
Show the answer
Answer: B.
$1 falls to roughly $0.624 across the three years compounded, a cumulative decline of about 38 percent, less than simply adding the three annual percentages.Aesop's formula for valuing an asset asks three questions about the birds in the bush. Which set is it?
- Revenue, users and market size
- How certain the birds are, when and how many will emerge, and the risk-free rate
- Price, volume and analyst coverage
Show the answer
Answer: B.
Buffett's 2000 letter frames every asset's value around those three questions, none of which is answered by a company's revenue growth alone.Sources
- Shiller, Irrational Exuberance, 2000, and the cyclically adjusted price to earnings ratio developed with John Campbell.
- Marks, memo "bubble.com", Oaktree Capital, January 2000.
- Berkshire Hathaway shareholder letter for 2000, written February 2001, on Aesop's bird in the hand, the three questions every valuation asks, and the Paine Webber-Gallup survey of investors, December 1999.
- Berkshire Hathaway 2025 annual report, table "Berkshire's Performance vs. the S&P 500": S&P 500 with dividends, by calendar year, 1965 to 2025.
What this leaves out
- The 100 times earnings example uses round, invented figures to show the arithmetic, not a record of any real company's multiple during this period.
- The shortcut assumes every dollar of profit is paid out; a company that must reinvest to grow needs more growth than it shows.
- Nothing here is advice about any security, and nothing here says a high multiple is always wrong.