IPO Arc research
IPO lock-up expiries 2014–2025: a slow drift, no clear drop on the day
What US IPO stocks did from 20 sessions before to 40 sessions after the lock-up expiry, 2014–2025
- Published
- Cohort
- US operating-company IPOs priced 2014–2025, all deal sizes, SPACs excluded; Yahoo Finance daily closes through 2026-11-06
Descriptive historical research: it describes how prices and volume behaved, not what any investor ought to do.
Summary
In 1,187 US IPOs priced 2014–2025 (prices through 2026-11-06), prices drifted lower around the lock-up expiry, but no faster than in the months before, and the expiry days showed no clear move. This runs against the common belief that the expiry itself pushes prices down. It does not rule out a small effect: against comparable IPOs and against the same IPOs' neighbouring months, the 60-session difference ranged from −1.9% to +3.2% depending on the comparison, and from −4.1% to +0.9% for deals of $50M and over (−4.1% against older peers, 95% CI −7.1% to −0.1%); for IPOs priced 2014–2020 it was −5.7% against older peers (95% CI −9.0% to −0.7%; 6 of 7 years negative), against +1.3% for 2021–2025; and $50M to $250M deals moved −0.6% on the expiry days (95% CI −1.3% to −0.1%).
- From 20 sessions before to 40 after the lock-up expiry, the median IPO returned −10.0% versus the S&P 500, a drift no faster than in the 60 sessions before (1,187 US IPOs priced 2014–2025, prices through 2026-11-06).
- The expiry days themselves (sessions −1 to +1) came to −0.1% versus the S&P 500, with a 95% interval of −0.5% to +0.3% (1,187 US IPOs priced 2014–2025, prices through 2026-11-06).
- Against comparable IPOs the 60-session difference ranged from −1.9% to +3.2% depending on the comparison, and from −4.1% to +0.9% for deals of $50M and over; against older peers it was −5.7% for IPOs priced 2014–2020 and +1.3% for 2021–2025 (1,187 US IPOs priced 2014–2025, prices through 2026-11-06).
- Volume on the expiry day was 1.47× a typical session for deals of $50M and over and 0.99× under $50M; $50M to $250M deals also moved −0.6% on the expiry days, with a 95% interval of −1.3% to −0.1% (1,170 US IPOs priced 2014–2025, prices through 2026-11-06).
- IPOs with an overhang of 8 or more shares not sold per share sold had a median of −18.8% versus the S&P 500 over the window, against −4.0% under 2, and part of that gap was there before the window (1,129 US IPOs priced 2014–2025, prices through 2026-11-06).
Limitations
- Survivorship: half the calendar has no usable free price history, and every usable company was still listed when the prices were collected, so the results describe survivors.
- Peers are not a perfect control: they share the pricing year, size and dates but not the age, and IPOs of the same age are usually at their own expiry. Older and younger peers bracket the comparison; the within-IPO step and the expiry days do not depend on peers.
- Overlapping events: many expiries fall in the same weeks (2020 and 2021 above all); the month-cluster bootstrap widens the intervals for this but does not remove the dependence between months.
- Dates: lock-up lengths come from automated (AI) extraction from the prospectus; 41 dates are estimated at 180 days. Some lock-ups count from the closing date, a few days after pricing; the three-session expiry window absorbs most of that.
- Early releases: the dates on which early releases took effect are being re-dated from the prospectus wording (earliest day, quarter and blackout rules); this study measures IPOs with such clauses only at their standard expiry, and a follow-up will measure the early dates.
- Overhang counts all shares not sold in the IPO, not only locked ones.
- Subgroups: size groups, overhang buckets, pricing halves and the 2014–2020 period split (chosen after a first look at the data) are several cuts of one sample; one of them can look unusual by chance.
- No costs and no rule: returns are close to close and before costs; the study fits nothing and tests no strategy.
Price path around the lock-up expiry
- All sizes
- Under $50M
- $50M–$250M
- Over $250M
- Middle 50%, all sizes
Data table
| Session | All sizes | Under $50M | $50M–$250M | Over $250M | 25th to 75th pct, all sizes |
|---|---|---|---|---|---|
| −20 | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% to 0.0% |
| −15 | −1.5% | −1.9% | −1.5% | −1.6% | −7.1% to +3.3% |
| −10 | −2.8% | −2.7% | −2.7% | −3.2% | −10.1% to +4.7% |
| −5 | −3.8% | −3.5% | −3.9% | −3.9% | −14.3% to +5.3% |
| −2 | −3.8% | −3.8% | −3.3% | −5.0% | −15.5% to +6.3% |
| −1 | −4.1% | −3.2% | −3.6% | −5.5% | −15.9% to +6.1% |
| 0(lock-up expiry) | −3.8% | −3.6% | −3.7% | −4.6% | −15.9% to +6.8% |
| +1 | −4.5% | −5.4% | −4.5% | −4.1% | −15.9% to +7.1% |
| +2 | −4.7% | −5.0% | −4.3% | −4.9% | −16.4% to +7.3% |
| +5 | −4.5% | −4.6% | −5.3% | −3.8% | −17.1% to +7.8% |
| +10 | −5.2% | −6.0% | −6.3% | −3.7% | −19.0% to +7.8% |
| +20 | −6.8% | −9.0% | −7.4% | −4.3% | −22.8% to +8.6% |
| +30 | −7.6% | −13.1% | −5.8% | −4.5% | −25.4% to +9.3% |
| +40 | −10.0% | −15.1% | −8.1% | −6.0% | −28.6% to +9.5% |
- Minus the S&P 500
- Minus older same-year peers
| Window | Minus the S&P 500 | Share negative | Minus older same-year peers |
|---|---|---|---|
| Run-insessions −20 to −2 | −3.8%−5.8% to −2.4% · n=1,187 | 61% | −1.7%−3.3% to +0.3% · n=835 |
| Expiry dayssessions −1, 0, +1 | −0.1%−0.5% to +0.3% · n=1,187 | 51% | +0.7%−0.04% to +1.4% · n=835 |
| Aftersessions +1 to +40 | −6.3%−8.7% to −4.2% · n=1,187 | 61% | +0.1%−2.7% to +1.8% · n=835 |
| Whole windowsessions −20 to +40 | −10.0%−13.0% to −6.6% · n=1,187 | 64% | −1.9%−4.6% to +2.0% · n=835 |
Expiry days (sessions −1, 0, +1), all sizes: mean −0.1% minus the S&P 500 (95% interval −0.9% to +0.7%). Volume on session 0: 1.35× a typical session of the 20 before the expiry (95% interval 1.26× to 1.45×), n=1,170.
- Session 0
- Session 0 is the first session on or after the lock-up expiry date (pricing date plus the prospectus lock-up length, or plus 180 days when estimated).
- Return
- Cumulative close-to-close return from the close of session −20.
- Minus the S&P 500
- Return minus the SPY return over the same dates.
- Minus older peers
- Return minus the median return of older IPOs of the same pricing year and size group over the same dates, leaving out peers at their own expiry.
- Volume
- Volume divided by the median daily volume of sessions −20 to −1.
- 95% interval
- Bootstrap 95% interval resampling calendar months of expiries.
Against comparable IPOs
Median difference over sessions −20 to +40 (95% CI; n) for each peer set, and the within-IPO step, by subset. Younger peers are still in their steeper early decline, which flatters the comparison; older peers are past theirs, which does the opposite. Peers within 30 sessions of age exist for only 144 IPOs (median +5.2%, 95% CI −6.1% to +10.5%): a peer of the same age is at its own expiry too unless its lock-up length differs.
- Measures. Raw return; return minus SPY over the same dates; and return minus the median of peers over the same dates. Peers are IPOs priced in the same calendar year, in the same size group, at least 20 sessions old at the window start, with prices through its end, and whose own −20 to +40 expiry window does not overlap it. IPOs of the same age are usually at their own expiry, so the peers are either older (priced earlier, past their own expiry) or younger. The main peer measure uses older peers (at least 3; 835 IPOs have them); section 4 also shows younger peers, an age-balanced mix (the average of the older and the younger median, at least 2 of each), all eligible peers and peers within 30 sessions of age.
- Within-IPO step. The expiry window (sessions −20 to +40) minus the average of the same IPO's 60 sessions before it (−80 to −20) and after it (+40 to +100), all minus SPY. It compares each IPO with itself at neighbouring ages.
| IPOs | Younger share of peers | Older same-year peers | Age-balanced peers | All eligible peers | Younger same-year peers | Same IPO, step |
|---|---|---|---|---|---|---|
| all sizesn=1,187 | 25% | −1.9%−4.6% to +2.0% · n=835 | +1.1%−3.6% to +5.2% · n=484 | −0.2%−2.5% to +2.1% · n=1,089 | +3.2%+0.02% to +7.6% · n=623 | −1.8%−4.7% to +0.9% · n=1,170 |
| $50M and overn=803 | 22% | −4.1%−7.1% to −0.1% · n=537 | −3.5%−6.3% to +1.7% · n=247 | −1.4%−4.1% to +1.2% · n=730 | +0.9%−2.9% to +4.4% · n=367 | −2.6%−5.2% to +0.3% · n=790 |
| under $50Mn=384 | 28% | +3.6%−4.2% to +7.5% · n=298 | +6.7%−0.4% to +11.3% · n=237 | +3.6%−2.2% to +8.3% · n=359 | +11.1%+3.1% to +15.3% · n=256 | 0.0%−6.2% to +5.2% · n=380 |
| $50M to $250Mn=458 | 21% | −6.2%−10.5% to +0.5% · n=319 | −8.0%−12.0% to −4.1% · n=138 | −4.9%−7.7% to −0.01% · n=429 | −5.5%−7.9% to +0.4% · n=215 | −5.0%−7.5% to −2.0% · n=449 |
| over $250Mn=345 | 24% | −2.3%−5.0% to +3.3% · n=218 | +5.8%−3.5% to +10.3% · n=109 | +1.8%−2.4% to +5.4% · n=301 | +8.5%+1.8% to +14.3% · n=152 | +0.1%−3.9% to +3.9% · n=341 |
| priced Jan–Junn=574 | 62% | +2.5%−4.6% to +6.9% · n=245 | +4.5%−0.3% to +9.3% · n=237 | +3.1%−0.1% to +6.8% · n=483 | +4.4%+0.5% to +8.7% · n=421 | −1.0%−4.1% to +3.4% · n=568 |
| priced Jul–Decn=613 | 10% | −3.1%−7.1% to +0.2% · n=590 | −2.1%−8.7% to +3.1% · n=247 | −3.3%−6.7% to −0.3% · n=606 | +0.5%−5.8% to +10.7% · n=202 | −3.6%−6.7% to +1.2% · n=602 |
| 2014–2020n=532 | 21% | −5.7%−9.0% to −0.7% · n=331 | −8.2%−12.0% to −4.3% · n=138 | −5.0%−7.1% to −1.6% · n=464 | −4.3%−7.7% to +0.1% · n=206 | −4.9%−7.5% to −0.8% · n=525 |
| 2021–2025n=655 | 26% | +1.3%−3.0% to +5.8% · n=504 | +6.5%+2.5% to +10.2% · n=346 | +4.2%−0.1% to +6.8% · n=625 | +9.3%+4.4% to +12.9% · n=417 | +1.0%−2.1% to +5.3% · n=645 |
| $50M and over, 2014–2020n=451 | 21% | −5.5%−8.8% to +0.5% · n=290 | −7.7%−12.0% to −4.3% · n=119 | −4.4%−6.7% to −0.6% · n=408 | −4.7%−7.9% to −0.01% · n=186 | −4.7%−7.9% to −0.2% · n=445 |
| $50M and over, 2021–2025n=352 | 24% | −3.0%−5.8% to +4.5% · n=247 | +5.4%−1.2% to +13.9% · n=128 | +2.2%−1.7% to +7.3% · n=322 | +7.9%+3.1% to +11.6% · n=181 | −0.7%−3.7% to +4.4% · n=345 |
How far to trust these numbers
How far to trust these numbers. Only 1,187 of the 2,368 operating-company IPOs in the Nasdaq calendar for 2014–2025 (50%) have usable free price histories, and all of them are companies that were still listed when the prices were collected. Companies that were delisted or acquired are missing, so the absolute and index-relative declines are probably understated; the direction of any bias in the peer comparison is unclear. Expiries cluster in busy IPO months, so the intervals resample whole months.
| Deal size | IPOs in calendar | Usable prices | No price history | Ticker reused or mismatched | Broken series | Coverage | In the study |
|---|---|---|---|---|---|---|---|
| Under $50M | 828 | 384 | 288 | 56 | 100 | 46% | 384 |
| $50M–$250M | 970 | 458 | 433 | 50 | 29 | 47% | 458 |
| Over $250M | 570 | 345 | 192 | 31 | 2 | 61% | 345 |
| All sizes | 2,368 | 1,187 | 913 | 137 | 131 | 50% | 1,187 |
Full report with every table (by year, deal size, overhang) by email
This box signs up for IPO Arc Weekly, one email a week about US IPOs. The confirmation email includes the link to the full report (the results by IPO year, deal size, overhang and early-release clause, the neighbouring windows, volume by session and the full method); the link works without confirming. IPO Arc Weekly starts only after the subscription is confirmed. Free; only an email address is needed. Past issues
Sources
Nasdaq IPO calendar; Yahoo Finance daily prices and volume, cut at 2026-11-06; lock-up terms from final prospectuses (424B4) and earnings dates from 8-K filings on SEC EDGAR. Yahoo Finance revises adjusted history and drops delisted tickers, so a later run can differ slightly.