Dotcom bubble & bust (1998-2002)

Definition

The Nasdaq rose ~4x from late 1998 to March 2000, then fell ~78% to its 2002 low; hundreds of internet companies priced on “eyeballs” and revenue multiples went to zero. The episode is the KB’s canonical VALUATION-DISCIPLINE case study — and, through Brunnermeier-Nagel’s finding that sophisticated investors rode the bubble rather than fighting it, its canonical lesson on the limits and timing costs of being right early.

How it works / structure

• The valuation record (Ofek-Richardson): aggregate internet-sector pricing at the peak implied growth assumptions arithmetically impossible for the sector as a whole (fa-dcf-valuation reverse-form logic — the implied paths were the falsifier, and they failed); short-sale constraints (thin borrow, lockup-restricted float — ms-short-locate-borrow, event-ipo-lockups) let optimists set prices without the pessimists’ capital.
• The unlock mechanics: Ofek-Richardson tie the collapse’s timing partly to lockup expirations flooding float — supply arrived as insider selling met exhausted demand.
• The smart-money finding (Brunnermeier-Nagel): hedge funds were OVERWEIGHT tech on the way up and cut before the collapse — riding, not correcting, the mispricing; shorting early was documented capital destruction even when the thesis was right (strategy-short-selling timing discipline; “the market can stay irrational” is this episode’s arithmetic).
• Engine-relevant fingerprints: extreme valuation dispersion, IPO/lockup calendars as supply schedules, retail attention concentration (sent-news-social pre-social-media form), and new-metric proliferation (“eyeballs” — metrics invented to justify prices, fa-sector-metrics non-GAAP discipline).

When it applies

Cited when implied-growth arithmetic is the falsifier; when short constraints explain persistent overpricing; when “right early is wrong” timing discipline needs its exhibit; when new-era metrics replace cash economics in a sector’s self-description.

Risk profile & failure modes

• The valuation lesson: aggregate implied growth is checkable arithmetic; sectors CAN price impossibility for years when short supply is constrained.
• The timing lesson: fighting a crowded narrative without a supply/catalyst schedule (lockups were one) donates capital regardless of eventual vindication.
• The survivor’s caveat: the same period built the companies that dominated the next two decades — bubble pricing and real transformation coexisted; blanket cynicism was as wrong as blanket euphoria.
• Misuse: pattern-matching every hot sector to 2000 without the borrow/lockup/implied-growth checks that made the diagnosis rigorous.

Evidence & limits

Ofek-Richardson (2003) and Brunnermeier-Nagel (2004) are peer-reviewed anchors; the price history is public record. Attribution of the collapse’s exact trigger remains partly narrative — the supply mechanics (lockups, issuance) are the best-documented component.

Falsifiable-thesis examples

Illustrations only, not signals:

• “Sector S’s aggregate price implies revenue growth above its addressable market’s arithmetic ceiling (Ofek-Richardson check)” — falsified by the implied-path computation.
• “Sector S’s insider-unlock calendar over the next two quarters exceeds 10% of current float (supply-schedule fragility)” — falsified by the filing arithmetic.

Cross-references

• The valuation machinery: fa-dcf-valuation (reverse form), fa-multiples-comparables, fa-sector-metrics
• The constraint mechanics: ms-short-locate-borrow, event-ipo-lockups, event-secondary-offerings
• The behavioral layer: bias-herding, sent-news-social
• The timing discipline: strategy-short-selling