The Dot-Com Bubble by the Numbers: Who Died, Who Survived, and How Long Recovery Took

Timeline of the dot-com bubble, 1995 to 2002

There is no official start date for a bubble, but most histories begin with a browser company that had been in business for 16 months and had never turned a profit.

1. August 9, 1995. Netscape goes public at $28 a share. It trades as high as $74.75 and closes its first day at $58.25, a valuation of nearly $3 billion.
2. December 5, 1996. Fed Chairman Alan Greenspan, in a speech at the American Enterprise Institute, asks how anyone can know when “irrational exuberance” has pushed asset prices too high. The S&P 500 closed that day at 744. It would more than double before it peaked.
3. November 13, 1998. theGlobe.com prices its IPO at $9 and closes the first day at $63.50, about 606% higher, a record first-day gain at the time.
4. May and November 1999. eToys and Webvan go public. Each ends its first day of trading valued at close to $8 billion.
5. March 10, 2000. The Nasdaq Composite closes at 5,048.62. It will not close higher for 15 years.
6. March 24, 2000. The S&P 500 closes at 1,527.46, its bubble-era high. Within days Cisco passes Microsoft as the most valuable listed company.
7. November 7, 2000. Pets.com announces it will shut down, nine months after its IPO.
8. March 2001. The US economy tips into recession, ending a 120-month expansion, the longest on record at the time.
9. July 9, 2001. Webvan stops operating and lays off about 2,000 people.
10. October 9, 2002. The bottom: Nasdaq 1,114.11, S&P 500 at 776.76.

How deep the Nasdaq 2000 crash went

Close to close, the Nasdaq Composite lost 77.9% between March 10, 2000 and October 9, 2002. The Nasdaq-100 tracker QQQ, which had launched only on March 10, 1999 and gained 131% in just over a year, fell 83.0% from its March 27, 2000 peak to the October 2002 low, even with dividends reinvested.

The S&P 500 price index, which excludes dividends, fell 49.1%, from 1,527.46 to 776.76. That sounds mild next to the Nasdaq until you look at what came before. On January 3, 1995 the index stood at 459. By March 2000 it had more than tripled. By the end of 2003, even after a 26% gain that year, it was still 27% below the peak.

The S&P 500 first closed above its 2000 high on May 30, 2007, more than seven years later. It did not stay there. The 2008 crisis dragged it down to 676.53 in March 2009, and the price index did not get above 1,527 for good until March 2013. The full story of that stretch is in our article on the lost decade of 2000 to 2009.

Companies that survived the dot-com bubble: the drawdowns

The table below covers the best-known tech companies that came through the bust intact. For each one we took its highest close in 1999-2001, the lowest close of the bust that followed, and the first day it closed back above that peak. Stock and QQQ figures are adjusted for splits and dividends, so the recovery dates reflect a holder who reinvested every dividend.

Three things stand out. First, nobody was spared: the mildest decline among these names was Microsoft's 65%, and that was a hugely profitable company with a near-monopoly in PC operating systems. Second, depth did not predict recovery time. Amazon fell the hardest of the group, 94%, yet regained its peak in under ten years, while Cisco, Intel and Qualcomm needed 14 to 21. Third, the price you paid mattered as much as the business. Buyers of Cisco at the top in March 2000 have only about doubled their money in 26 years, dividends included. Buyers of Amazon at its December 1999 top have about 46 times theirs.

The table also shows only the names that are still around, which flatters the era. Lists like this one are a textbook case of survivorship bias: an index of 2000-vintage internet stocks picked today automatically leaves out the ones that went to zero.

Cisco: the most valuable company in the world, then a 21-year wait

Cisco sold the routers and switches the internet ran on, and in early 2000 investors priced it as if that demand would compound forever. At the end of March 2000 its market value passed $500 billion and it overtook Microsoft as the most valuable listed company. Its share price closed at a split-adjusted $80.06 on March 27, 2000.

Then demand turned. Cisco had ramped up production for growth that did not arrive, and in 2001, in the networking industry's first steep downturn, it wrote down $2.25 billion of inventory. By October 8, 2002 the stock had lost 89% of its value. From such a high starting price, the shares, with dividends reinvested, only regained the 2000 peak in August 2021. The share price itself closed at a record $80.25 on December 10, 2025, more than 25 years after the old high. Comparisons between Cisco then and today's AI chip leaders are covered in AI boom vs dot-com bubble.

Dot-com bubble companies that failed

The dead are better remembered than many of the survivors, partly because their rise and fall were so compressed.

• theGlobe.com, an early online community, set the first-day IPO record in November 1998 when its $9 shares closed at $63.50 after touching $97. In August 2001 it closed its flagship site and laid off half its staff.
• eToys priced its IPO at $20 in May 1999 and closed the first day around $76, which gave it a higher market value ($7.8 billion) than Toys “R” Us. Holiday sales in 2000 fell far short of forecasts, the shares dropped below $1 before the year was out, and in early 2001 the company shut down and went into Chapter 11.
• Webvan, the online grocer, went public at $15 in November 1999 and ended its first day at $24.875, worth $7.9 billion. It had lost about $830 million by July 9, 2001, when it shut down with shares at 6 cents.
• Pets.com sold shares at $11 in February 2000 and announced its liquidation on November 7 of the same year. The full case is in Pets.com: from IPO to liquidation in 268 days.

The bust also exposed frauds that had little to do with selling things on the web. Enron filed for bankruptcy in December 2001 and WorldCom, the telecom giant, in July 2002. Both are covered separately: the Enron collapse and the WorldCom accounting fraud.

Why winners and losers looked alike in real time

Looking back, Amazon and Pets.com seem like different species. In 2000 they had a lot in common: both sold goods online, both lost money on heavy spending, and both stocks were falling hard. Barron's ran a cover story in May 1999 that called Amazon “Amazon.bomb”, and in March 2000 the same paper published a widely read tally of internet companies that were running out of cash.

In June 2000 Ravi Suria, a convertible-bond analyst at Lehman Brothers, argued that Amazon could run out of cash within four quarters. The stock fell about 20% on June 23. He was wrong, but not for reasons a stock chart would have shown. Amazon had nearly $1 billion in cash and securities, partly from $672 million of convertible bonds it had sold to European investors in February 2000, just before that window shut. It reported its first quarterly net profit, $5 million, for the fourth quarter of 2001, and its first full-year profit, $35 million on $5.26 billion of sales, in 2003.

The price action gave no help at all. Amazon fell further than eBay or Cisco, and at its September 2001 low, down 94%, its chart was hard to tell apart from that of a retailer heading for zero. What separated the survivors was mostly in the filings: how much cash was in the bank, how fast it was burning, and whether the core business could eventually pay for itself. Even that was a judgment call, as the Lehman report shows.

What the Fed and the economy were doing

The Federal Reserve was raising rates into the top. The effective fed funds rate averaged 6.54% in July 2000, and unemployment hit 3.8% in April 2000. The Fed then cut aggressively: the rate was 1.73% by January 2002 and below 1% at the end of 2003. The recession itself was short, March to November 2001, but stocks kept falling for almost another year, and unemployment kept rising until it peaked at 6.3% in June 2003. Lower rates did not stop the repricing of stocks that had been valued for flawless growth.

How to play the dot-com era blind

Every number above was computed with hindsight. The harder exercise is to sit in 1998 or 2000 without knowing which companies will make it. That is what Blindfolio is built for. Pick the 1990s as your start decade and the game drops you into a 10-, 15- or 20-year window that begins somewhere in that decade. The year is shown on screen in this mode, but the game, not you, decides when the window starts, so the bust may arrive early, late or not at all. Leave the start on random and the years are masked too.

Time moves forward a week at a time. You get a monthly income to invest, a deposit that pays the Fed funds rate of the day, and real news headlines from the period. Companies trade under codenames from the point they historically went public, and the catalog includes failures such as Pets.com and Webvan alongside the survivors. At the end the real names are revealed and your portfolio is compared with the S&P 500, dividends reinvested, over the same years. For a class, everyone can pick the 1990s and the same round length: each student lands in a different window inside the decade, and comparing who caught the crash and who missed it makes a good discussion.