IPO lockups

Definition

IPO lockup agreements bar insiders and pre-IPO holders from selling for a set period after listing — typically 180 days, increasingly with staggered or price-conditioned early releases. Expiration is a SCHEDULED SUPPLY EVENT: the float can multiply overnight, and the documented record shows abnormal negative returns and volume spikes around the date — one of the few calendar events with a persistent (if modest) peer-reviewed directional footprint.

How it works / structure

• The mechanics: lockup terms live in the prospectus (S-1/424B — dates, share counts, early-release triggers); the locked shares often dwarf the IPO float (a 10-15% float at listing means 85-90% unlocks later); expiration converts scarcity pricing into supply pricing.
• The evidence: Field-Hanka (2001) documented a statistically significant ~1.5% abnormal return drop at expiration with a 40% permanent volume increase, larger for venture-backed firms — an anomaly they note is not arbitrageable at scale because of shorting costs into the event (ms-short-locate-borrow — pre-unlock borrow is scarce and expensive precisely because the trade is obvious).
• Engine-executable structure: lockup calendar per name (dates and share counts from filings), float-multiple at unlock (shares releasing ÷ current float), borrow-fee series into the event, staggered/early-release conditions (price triggers change the date).
• The modern wrinkles: staggered unlocks, early releases on strong prices, and direct listings without lockups — per-name terms have diversified; the calendar is filing work, not a formula.

When it applies

Recent-IPO positioning (any thesis on a name inside its first year checks the unlock calendar); supply-pressure theses at expiration (sized against the float multiple and the borrow reality); insider-behavior reading after the unlock (sent-insider-transactions — post-lockup sales are expected; their SIZE is the signal).

Risk profile & failure modes

• The obvious-trade tax: shorting into unlocks pays spiking borrow fees that consumed the documented edge in the original study — the anomaly persists partly BECAUSE it resists arbitrage.
• Pre-positioning reversal: unlock-day selling by front-runners covers into the event — the price path around the date is two crowds passing.
• Early-release surprises: price-triggered early unlocks move the date; stale calendars trade the wrong day.
• Overhang vs event: names can bleed for weeks BEFORE the date (anticipated supply) — the event window is wide and the timing imprecise.

Evidence & limits

Lockup terms are SEC-filed facts. Field-Hanka (2001) and successor studies document the negative abnormal return and volume shift; the shorting-cost barrier to arbitrage is part of the documented finding. Effect sizes are modest and per-name variance is large — cohort theses, not single-name certainties.

Falsifiable-thesis examples

Illustrations only, not signals:

• “X will underperform its sector in the two weeks spanning its lockup expiration (float multiple 4x, borrow under 15%)” — falsified by the realized relative return.
• “X’s borrow fee will exceed 25% annualized in the week before the unlock (crowding gauge)” — falsified by the fee series.

Cross-references

• The barrier that preserves it: ms-short-locate-borrow
• The aftermath read: sent-insider-transactions
• Filing sources: fa-guidance-estimates (prospectus/filing discipline)
• Float mechanics: ms-liquidity