Party Like It's 1999?

1. Late to the Party

Thiel opens with a generational point: someone who was a toddler in 1969 can read about civil rights and Vietnam, but never truly feels those debates the way people who lived them do. The 1990s play the same role for technology. Students in 2012 were children during the boom, yet the entire landscape they operate in — which ideas get funded, which behaviors are taboo — was cast in that fire. So before asking whether startups make sense now, you have to reconstruct what the decade actually felt like.

It is questionable whether one can really understand startups without, say, knowing about Webvan or recognizing the Pets.com mascot.

2. A Quick History of the '90s

The euphoria after the Berlin Wall fell in November 1989 was brief. From 1990 to about 1994 the U.S. slogged through recession: manufacturing kept declining, the shift to a service economy was slow and painful, and the culture turned pessimistic — think Nirvana and grunge. Ross Perot mounted a serious third-party run, and George H.W. Bush became a one-term president. In the late 1980s, Japan looked set to dominate semiconductors, and even at Stanford the tech world felt distant.

The Internet changed the mood. Xanadu had imagined a two-way network between all computers back in 1963, but it needed everyone to adopt it at once; it raised venture money for some 29 years without shipping and finally died in 1992. Netscape arrived in 1993 with a workable server-client model, and its August 1995 IPO woke the public up: priced at $14, doubled before trading, doubled again on day one, and hit $160 within five months — for an unprofitable company. Hubris followed: Netscape taunted Microsoft on its own campus, Bill Gates threw the company at the Internet, IE ate Netscape's share, and the pioneer ultimately sold to AOL for over a billion dollars largely on the value of its antitrust claims. The next three years were quiet: by late 1998 the NASDAQ sat near 1,400, only 400 points above August 1995. Yahoo went public in 1996 at a $350M valuation, Amazon in 1997 at $460M, and skeptics scoffed at the multiples.

• December 1996: Alan Greenspan warns of 'irrational exuberance' — roughly three years early.
• 1997: The East Asian financial crisis flattens Thailand, Indonesia, South Korea, and Taiwan.
• 1998: Russia's ruble crisis, then the leveraged collapse of Long-Term Capital Management, contained only by a Fed-orchestrated bailout.
• January 1999: The euro launches to immediate skepticism and promptly loses value.

Put together, the mania was a proof by elimination. The old economy could not compete with cheap labor in Mexico and China. Emerging markets had just revealed themselves as crony capitalism. Europe inspired little confidence, and after LTCM nobody trusted leverage. With every alternative discredited, money defaulted into the one thing left: technology.

3. The Mania: September 1998 – March 2000

The dot-com mania was intense but shorter than people remember: roughly September 1998 to March 2000. Money was everywhere, and so were sketchy operators. Launch parties happened nightly, ranked by an exclusive email list. Forty-year-old grad students ran multiple dubious companies at once; a billionaire from Idaho handed capital to anyone with a polished pitch; broke founders picked up thousand-dollar dinner tabs and paid in shares. Business models went negative-margin — losing money on every customer while vowing to make it up in volume — and merely adding '.com' to a name could roughly double a company's value. Yahoo, by then the Valley's largest Internet company, justified stock-for-stock acquisitions with the claim that its stock only went up.

1. 1

   VA Linux: a billion for a day

   Larry Augustin nearly shut VA Linux in 1997 but kept going. Its 1999 IPO priced at $30, traded up to $300 the same day — the biggest first-day pop ever — making his 10% stake worth about $1 billion. By the end of the lock-up six months later the stock had lost 90%, then another 90% over the next six. He walked away with $5–6 million. He had also once declined to be Yahoo's third employee.

2. 2

   The aura test

   With sketchy people everywhere, Max Levchin developed a filter: size up anyone pitching you within about 15 seconds, and if the aura is off, walk away. Crude, but companies that screened hard survived the flood of bad actors better than those that did not.

3. 3

   Negative-margin economics

   The era's signature absurdity: business models that lost more per customer than they earned, like a bank paying more to sort $100 in pennies than the deposit is worth. Growth metrics masked the fact that scaling only scaled the losses.

PayPal's wild ride

PayPal started in December 1998 and deliberately hired only friends — a defense against the era's sketchy talent pool. The original idea, beaming money between Palm Pilots, was voted one of the ten worst business ideas of 1999. Angel investors barely cared what the product was; one asked only who else was investing, then reportedly consulted a fortune cookie. Nokia Ventures put in $4.5 million, and at the very first board meeting after the investment, the team announced a pivot: mobile infrastructure was years away, so PayPal became an account system for emailing money to anyone.

Growth was the next crisis. Advertising was too expensive and business development with big banks went nowhere — a meeting Luke Nosek arranged with HSBC executives convinced the team BD was hopeless. So PayPal bought virality directly: $10 for signing up, $10 per referral. The user base grew 7–10% per day, but each customer cost about $20 and revenue was zero, so costs grew exponentially too. The company needed buzz to raise more money to keep going. A flattering Wall Street Journal piece on February 16, 2000 tossed off a back-of-envelope $500M valuation; the lead investor in the next round treated that number as authoritative. A South Korean firm wired $5 million without documents or negotiation and refused every attempt to return it. PayPal closed $100 million on March 31, 2000 — days after the market peaked. The luck of that timing funded its survival; Thiel notes this worked, but is hardly a recommended way to run a company.

4. Hubris and Schadenfreude

Prince's song turned out to be literal: the party ended right on schedule. In the first half of 2000, marketing-driven e-commerce companies died; in the second half, the B2B companies followed; in 2001, telecom collapsed. Thiel adds a telling inversion: in March 2000, arguably the single most depressed sector was military defense — with the NASDAQ soaring, nobody believed there would ever be another war — and defense then rose for most of the following decade. Peak optimism about one future was peak blindness about another.

Culturally, enormous hubris gave way to schadenfreude: the skeptics declared they had been right all along, and the mood curdled into depression. PayPal, meanwhile, ground its way to breakeven in 2001 by solving fraud and customer service, and in late September 2001 became the first company to file for an IPO after 9/11. Twenty months after the rosy profile, the Wall Street Journal ran a piece titled 'Earth to Palo Alto,' sneering at a three-year-old company that had never turned an annual profit, was on track to lose a quarter billion dollars, and whose own filings warned of fraud and money-laundering risks — yet whose managers and VCs were taking it public. The kicker was brutal.

The U.S. needs [PayPal] as much as it does an anthrax epidemic.

5. Lessons Learned — and Overlearned

The world's post-2000 narrative: the bubble was pure destruction, so return to the real economy — 'bricks and clicks' became clicks going back to bricks. Money rotated into housing and emerging markets; Warren Buffett's old-economy stance looked vindicated; only profits could justify valuations; globalization beat technology; the future was declared fundamentally unknowable, all prophets false, all claims suspect. Thiel's critique is that these lessons were driven by hubris, envy, and resentment — emotions that make for bad analysis. People in the 1990s were right about a great deal: the euro really was shaky, crony capitalism and overleverage really were problems, and the grand belief in technology was directionally sound even when the prices were insane.

March of 2000 wasn't just a peak of insanity. In some important ways, it was still a peak of clarity as well.

6. Bubbles: Are We in One?

Is 2012 another bubble? There are frothy data points: more Stanford students studying computer science than in 1999, valuations creeping upward. But scattered froth does not make a bubble. A bubble requires two things: widespread, intense belief, and that belief being false. Thiel's diagnosis is that society in 2012 no longer intensely believes in much of anything — so the precondition fails. The insistent bubble narrative comes from people hunting for one, which is itself an overreaction to the pain of the 1990s rather than good analysis.

The tempting alternative — antibubble thinking, insisting everything will work and everyone should load up on houses and tech stocks — is probably somewhat closer to true, but Thiel rejects it too. Both positions make the same mistake: they treat truth as a social fact, something to be read off the crowd, whether by copying it or inverting it.

If the herd isn't thinking at all, being contrarian—doing the opposite of the herd—is just as random and useless.

The genuinely contrarian move is to think for yourself. Instead of asking whether there is a bubble, ask: is this specific company valuable? Why? How would you actually figure that out? Those questions have answers you can work toward — and they set up the rest of the course.

Then vs. now (2026)

Self-check quiz

Pick an answer to reveal the explanation.