A stock performs well for months after its IPO. Then it drops sharply. No obvious bad news appears. In many cases, the trigger is a lockup expiration.
The lockup expiration date is set and disclosed at the moment an IPO prices. It is not a surprise. Yet decade after decade, the data shows the market still reacts to it. That is the central puzzle this article examines.
Academic studies show the average abnormal return around lockup expiration is modest, typically in the −1.5% to −2% range. Individual high-growth, VC-backed companies with low float can experience far larger moves. This is especially true when the unlock coincides with weak earnings, valuation pressure, or concentrated insider selling.
What a Lockup Is and Why It Exists
When a company goes public, only a small portion of its total shares, usually 10% to 20%, are sold in the IPO itself. The rest stays in the hands of founders, employees, and early investors. A lockup agreement restricts those holders from selling for a fixed period, typically 180 calendar days.
The rationale is straightforward supply-and-demand logic. If every pre-IPO holder could sell on day one, the market would be flooded with stock immediately. The company would have little time to establish a trading history or report its first public earnings. The lockup buys that time.
When the lockup expires, the freely tradable float can triple or quadruple overnight. That is the structural event the research below measures.
How Researchers Measure It: Event-Study Methodology
To isolate the price effect of a lockup expiration, researchers use a method called an event study. The goal is to strip out everything unrelated to the lockup itself.
This is done by calculating abnormal returns. The stock's actual return during a specific window around the unlock date is measured. A comparable benchmark return, such as the Russell 2000 or a relevant IPO index, is then subtracted from that actual return. What remains is the portion of the move plausibly attributable to the lockup expiration alone.
Key Formulas
Abnormal Return
Abnormal Return = Stock Return − Benchmark Return
This isolates returns not explained by broader market movement.
Cumulative Abnormal Return (CAR)
CAR = Σ Abnormal Returns across the event window
This sums daily abnormal returns over the days before and after the lockup expiration, commonly a three-day or five-day window.
Float Increase at Unlock
Float Increase = Newly Unlocked Shares ÷ Pre-Expiration Public Float
This measures how sharply the tradable supply expands.
Unlock Ratio
Unlock Ratio = Shares Becoming Eligible for Sale ÷ Total Shares Outstanding
This shows the scale of potential selling relative to the full cap table.
What the Research Actually Shows
Two major academic studies anchor the empirical evidence here.
Field and Hanka (2001), published in the Journal of Finance, examined 1,948 US lockup agreements. They found a permanent 40% increase in average trading volume at expiration. They also documented a statistically significant three-day abnormal return of −1.5%. The effect was much larger when the company was venture capital-backed. VC firms sold more aggressively than executives or individual shareholders.
Brav and Gompers (2003), published in the Review of Financial Studies, examined 2,794 US IPOs. They found an abnormal return of roughly −2% clustered around the lockup expiration date. Their results also supported the interpretation that lockups serve as commitment devices. Insiders at firms with greater potential for moral hazard locked up their shares for longer.
Study | Sample Size | Key Finding |
Field and Hanka (2001) | 1,948 US lockup agreements | Permanent 40% volume increase; three-day CAR of −1.5% |
Brav and Gompers (2003) | 2,794 US IPOs | Abnormal return of roughly −2% around expiration date |
Two separate research teams, with different sample sizes, reached the same basic conclusion. Lockup expirations move prices, even though the date was never a secret.
The price reaction is also tightly clustered. Field and Hanka found the effect concentrated mostly within a three-day window around the unlock date. It does not spread gradually over weeks. It arrives at a known date and resolves quickly under normal conditions.
Why VC Backing and Sector Amplify the Effect
Not all lockup expirations are equal. The research consistently identifies two amplifiers.
Venture capital and private equity backing. During the IPO itself, VC involvement is generally a quality signal, sometimes called the certification effect. That relationship reverses at lockup expiration. VC firms operate on fixed fund timelines and answer to limited partners who require capital returned on a schedule. That structural pressure makes VC-driven selling more aggressive than selling by an individual founder with no fund clock running. Field and Hanka's 1,948-deal dataset confirmed VC-backed firms showed significantly larger abnormal volume and price reactions than non-VC-backed peers.
High-growth and technology-sector firms. Companies in fast-moving sectors, especially those that ran up significantly after their IPO, tend to see the sharpest reactions. These companies carry more inherent uncertainty. When insiders begin selling at the lockup date, the market reads that activity as a stronger signal about insider confidence than it would for a more predictable, stable business.
Both amplifiers compound each other. A VC-heavy cap table in a high-uncertainty sector, combined with a single lump-sum release, is the configuration most likely to produce a severe price reaction.
Two 2025 Case Studies: CoreWeave and Figma
Theory is one thing. Two 2025 unlock events offer a direct comparison of the patterns the academic research predicts.
CoreWeave: A Textbook Supply Shock
CoreWeave, a heavily VC-backed AI infrastructure company, priced its IPO on March 27, 2025. Its lockup expiration was accelerated to August 14, 2025, triggered two trading days after its Q2 2025 earnings announcement, per CNBC's reporting on the earnings and lockup timeline.
Checking CoreWeave's actual SEC filing for the lockup terms directly, the acceleration trigger is written in fairly plain language; it isn't buried in the dense boilerplate some lockup clauses use, which makes it one of the easier lockup provisions to actually verify yourself on EDGAR.
The earnings themselves missed expectations. CoreWeave reported a wider-than-expected adjusted loss of $0.27 per share against an expected $0.21, according to CNBC.
Over 80% of CoreWeave's Class A shares became eligible for sale simultaneously, as Yahoo Finance reported. The stock fell 33% in the week following expiration. From its mid-June peak, the stock declined 46%, with the company's market cap dropping from roughly $88 billion to $49 billion, according to the same Yahoo Finance report.
This is close to a worst-case version of the academic pattern: VC-heavy ownership, high-uncertainty sector, single lump-sum release, and an earnings miss timed alongside the unlock.
Figma: A Staggered Release
Figma took a different structural approach. Its lockup architecture included multiple release windows rather than a single expiration date.
The first early release was a performance-based trigger. When Figma's stock satisfied the early release condition, 25% of employee-held shares became eligible for sale at the open of trading on September 5, 2025, as Figma disclosed directly in its SEC filing dated September 3, 2025.
Trefis reported that Figma's stock fell roughly 14% to 15% in after-hours trading following the earnings announcement and unlock disclosure, compared to a broader intraday drop reported across financial media.
The result was a meaningful but contained reaction, considerably softer than CoreWeave's multi-day crash. Figma also entered an extended lockup agreement on August 30, 2025, covering approximately 54.1% of its outstanding Class A shares through August 31, 2026, as disclosed in the same SEC filing.
The side-by-side comparison is directly useful for institutional analysis. A staggered release structure that breaks float expansion across multiple dates measurably dampens the supply shock. That holds even when both companies are VC-backed, high-profile, and AI-adjacent.
SPAC Lockups: A Different Risk Profile
SPAC lockup structures differ materially from traditional IPO lockups.
Dimension | Traditional IPO Lockup | SPAC Lockup |
Typical length | 180 days | 6 months to several years, varies by deal |
Who is subject | Founders, employees, early investors | Sponsors, PIPE investors, target shareholders |
Key risk | Standard supply shock at a known date | Sponsors may profit even when the stock has declined |
SPAC sponsors often acquired founder shares at a very low cost basis, sometimes effectively at nominal cost. Their financial incentive to sell at expiration remains high even if the merged company's stock has weakened since closing. SPAC sponsor lockups, founder-share lockups, PIPE investor lockups, and target-shareholder lockups can all carry different lengths and terms within the same deal. Any analysis of SPAC lockup risk should separate these layers rather than treating the expiration as a single event.
What Institutional Investors Watch Before Unlock Dates
Several signals are worth tracking ahead of any lockup expiration.
Underwriter waivers. Lead underwriters can release lockup restrictions early, often to support a secondary stock offering. FINRA Rule 5131 requires the book-running lead manager to announce any impending lockup release or waiver at least two business days in advance through a major news service. A surprise early waiver can move a stock independently of the actual selling, since the market sometimes reads it as a signal of weaker insider confidence.
Float overhang. A single large locked shareholder who did not sell during the IPO can disrupt price discovery when it finally exits. One large seller offloading a concentrated position in a thin market can move the price more than diversified selling from many smaller holders combined.
Cap table composition going into the unlock. Knowing how VC-heavy a cap table is, and whether the company falls into a high-uncertainty sector, gives a practical early read on the likely severity of a lockup reaction. The CoreWeave-to-Figma spectrum is a useful reference range. CoreWeave represents the concentrated, single-event, VC-heavy configuration. Figma represents the staggered, partially mitigated version of the same structural risk.
Limits of the Evidence
The average abnormal returns documented by Field and Hanka (−1.5%) and Brav and Gompers (−2%) are statistically significant but modest in absolute terms. They represent averages across thousands of deals, including many where the lockup expiration produced little visible market reaction.
Individual outcomes depend heavily on deal-specific factors: the concentration of locked shares, the sector, whether the stock ran up significantly post-IPO, timing relative to earnings, and whether the release was structured as a single cliff or staggered across dates.
The event-study methodology itself also carries limits. It measures abnormal returns relative to a benchmark, which means it depends on the choice of benchmark and the definition of the event window. Different studies use slightly different windows and benchmarks, which contributes to variation in reported figures across the literature.
Frequently Asked Questions
Is the average abnormal return of −1.5% to −2% always what investors should expect? No. That is a sample average across thousands of deals. Individual outcomes vary significantly. VC-backed, high-uncertainty, low-float names with a single lump-sum release can see far larger reactions, as the CoreWeave case demonstrates.
Does the market not simply price in the lockup expiration in advance? Partially. Markets anticipate the event, which is why some pressure appears in the days before the unlock. But Field and Hanka's data shows the price reaction is still statistically significant at the expiration date itself, suggesting anticipation is incomplete.
Where is a company's lockup structure disclosed? It is disclosed in the S-1 registration statement before the IPO. Staggered or performance-based structures, like Figma's, are also detailed there. Comparing the original S-1 against later amendments on SEC EDGAR shows exactly how the structure is set up, well before expiration.
Can a lockup be waived or shortened after the IPO? Generally only with underwriter consent, which must then be publicly disclosed under FINRA Rule 5131. Companies cannot quietly change lockup terms after the fact without that disclosure trail.
Does a staggered lockup eliminate the price impact? No. It reduces the magnitude of any single supply shock by distributing float expansion across multiple dates. Figma's September 5, 2025 release still produced a 14% to 15% intraday drop. The impact is smaller than a single-event unlock, but it is not eliminated.
