IPO Performance Statistics 2026
By Axis Intelligence Research
Co-author: Mia Scarlett (Business) | Last updated: September 16, 2026 | License: CC BY 4.0
Buy the average U.S. IPO at its first closing price, hold it three years, and you earned 19.1% while the market beat you by 20.5 points. Buy the same stocks at the offer price and the shortfall narrows to 3.3 points. Axis Intelligence Research calls that 17.2-point difference the IPO Allocation Gap (IAG).
Quick Answer
- Do IPOs beat the market? Not for investors who buy on the open market. Across 9,253 U.S. operating-company IPOs from 1980 to 2024, the average three-year buy-and-hold return from the first close was 19.1%, or −20.5% after adjusting for the market, according to Jay R. Ritter’s long-run IPO statistics (University of Florida, returns through December 31, 2025).
- Who captures IPO returns? The allocated investor. Axis Intelligence Research finds an IPO Allocation Gap (IAG) of 17.2 percentage points for all IPOs and 28.3 points for tech IPOs.
- Are IPO withdrawals rising? In Hong Kong, yes: 210 listing applications lapsed, were rejected, returned or withdrawn from January to July 2026, against 131 that listed, per HKEX.
Key Findings
- Axis Intelligence Research finds that the IPO Allocation Gap (IAG) for U.S. IPOs from 1980 to 2024 is 17.2 percentage points of three-year market-adjusted return, rising to 28.3 points for tech IPOs.
- Axis Intelligence Research calculates that 60.0% of U.S. IPOs from 1975 to 2021 had a zero or negative three-year return from their first closing price, while 16.0% more than doubled.
- Axis Intelligence Research finds that 61.6% of Hong Kong IPO applications resolved between January and July 2026 dropped out rather than listed, up from 37.1% in 2025, based on HKEX monthly reports.
- According to Jay R. Ritter, the U.S. IPO classes of 2020, 2021, 2022 and 2023 trailed the market by between 68.6 and 88.3 percentage points over their first three years (or less, for 2023).
- Axis Intelligence Research calculates a Day-One Capture Ratio (DCR) of 52.1%: the average first-day return equals about half of the average three-year return measured from the offer price.
Do IPOs Outperform the Market After They Go Public?
On average, no. The 9,253 U.S. operating-company IPOs that Ritter tracks from 1980 to 2024 returned a mean 18.9% on day one, then 19.1% over the following three years. The CRSP value-weighted market returned 39.6% over the matching windows. That is a −20.5% market-adjusted return and −8.9% against style-matched firms of similar size and book-to-market ratio.
Ritter annualizes the market shortfall at 5.5% per year over an average holding period of 2.8 years. Delisted companies are held to their delisting date, so failures stay in the sample.
IPO long-run returns by era, 1980–2024
| IPO period | IPOs | Mean first-day return | 1-year return | 3-year return | Market-adjusted | Style-adjusted | Source |
|---|---|---|---|---|---|---|---|
| 1980–1989 | 2,047 | 7.2% | 3.4% | 22.5% | −22.6% | 2.2% | Jay R. Ritter, Table 19 |
| 1990–1998 | 3,616 | 14.8% | 14.0% | 39.7% | −21.0% | −0.1% | Jay R. Ritter, Table 19 |
| 1999–2000 | 856 | 64.6% | −11.2% | −53.1% | −31.8% | −58.9% | Jay R. Ritter, Table 19 |
| 2001–2010 | 1,010 | 11.6% | 6.7% | 17.1% | 2.2% | −7.6% | Jay R. Ritter, Table 19 |
| 2011–2024 | 1,724 | 23.0% | −1.7% | 9.1% | −24.7% | −16.4% | Jay R. Ritter, Table 19 |
| 1980–2024 | 9,253 | 18.9% | 5.6% | 19.1% | −20.5% | −8.9% | Jay R. Ritter, Table 19 |
The one decade that beat the market, 2001–2010, is also the one that followed a crash, when only the strongest issuers could list. The 2011–2024 era carries a larger first-day return (23.0%) than any era outside the dot-com years and a worse market-adjusted outcome (−24.7%) than the 1980s.
IPO returns by class year, 2016–2024
| IPO class | IPOs | First-day return | 1-year return | 3-year return | Market-adjusted | Style-adjusted | Source |
|---|---|---|---|---|---|---|---|
| 2016 | 75 | 14.5% | 23.3% | 70.5% | 29.5% | 27.0% | Jay R. Ritter, Table 19 |
| 2017 | 106 | 12.9% | 32.4% | 52.8% | 22.6% | 35.6% | Jay R. Ritter, Table 19 |
| 2018 | 134 | 18.6% | −6.8% | 79.1% | 23.4% | 55.8% | Jay R. Ritter, Table 19 |
| 2019 | 113 | 23.5% | 33.0% | 12.5% | −25.1% | −8.8% | Jay R. Ritter, Table 19 |
| 2020 | 165 | 41.6% | 9.9% | −48.1% | −78.6% | −75.7% | Jay R. Ritter, Table 19 |
| 2021 | 311 | 32.1% | −49.1% | −49.1% | −68.6% | −42.3% | Jay R. Ritter, Table 19 |
| 2022 | 38 | 48.9% | −27.7% | −29.5% | −72.2% | −69.9% | Jay R. Ritter, Table 19 |
| 2023* | 54 | 11.9% | −11.6% | −30.9% | −88.3% | −84.7% | Jay R. Ritter, Table 19 |
| 2024* | 72 | 15.4% | −17.4% | −4.7% | −33.2% | −12.3% | Jay R. Ritter, Table 19 |
*Returns run through December 31, 2025, so the 2023 and 2024 classes are measured over less than three years.
The 2020–2023 classes averaged a −76.9% market-adjusted three-year return, by Axis Intelligence Research’s simple average of the four cohort means. The 2016, 2017 and 2018 classes, by contrast, each beat the market by more than 20 points.
When IPOs underperform: the first two years
Ritter’s event-year table isolates where the damage happens. IPO firms from 1980 to 2024 returned 5.6% in year one against 11.4% for size-matched companies, a 5.8-point gap, and 5.8% in year two against 13.7%, a 7.9-point gap. By year four the IPOs were ahead, 20.5% versus 16.3%.
The first six months returned 4.9%. The second six months returned 0.0%. For IPOs from 2010 to 2024, the first-year shortfall against size-matched firms widens to 9.4 points.
Mia Scarlett: Treat the first-day return as the issuer’s cost of capital and the second year as the bill for the story. A class that trails the market by 68.6 points, as 2021 did, is usually a class whose first-close price had already discounted the growth. The returns table reads like a valuation table with a lag.
How Many IPOs Lose Money Over Three and Five Years?
Most of them. Axis Intelligence Research calculates that 5,516 of the 9,195 U.S. IPOs from 1975 to 2021, or 60.0%, had a zero or negative three-year return measured from the first close. The median three-year return was −25.7% while the mean was 21.2%: a handful of very large winners pulls the average up.
Distribution of three-year IPO returns, 1975–2021
| 3-year buy-and-hold return | Share of IPOs, from first close | Share of IPOs, from offer price | IPOs (from first close) | Source |
|---|---|---|---|---|
| Below −50% | 38.5% | 35.0% | 3,537 | Jay R. Ritter, Table 16e |
| −50% to 0% | 21.5% | 21.1% | 1,979 | Jay R. Ritter, Table 16e |
| 0% to 50% | 14.5% | 14.7% | 1,333 | Jay R. Ritter, Table 16e |
| 50% to 100% | 9.5% | 9.9% | 873 | Jay R. Ritter, Table 16e |
| 100% to 200% | 8.3% | 9.6% | 759 | Jay R. Ritter, Table 16e |
| 200% to 500% | 6.0% | 7.4% | 552 | Jay R. Ritter, Table 16e |
| 500% to 1,000% | 1.3% | 1.7% | 123 | Jay R. Ritter, Table 16e |
| 1,000% to 2,000% | 0.3% | 0.5% | 27 | Jay R. Ritter, Table 16e |
| 2,000% to 3,000% | 0.1% | 0.1% | 6 | Jay R. Ritter, Table 16e |
| Above 3,000% | 0.1% | 0.1% | 6 | Jay R. Ritter, Table 16e |
Axis Intelligence Research calculates that 16.0% of IPOs (1,473 companies) more than doubled over three years from the first close. Measured from the offer price, 56.1% still ended at zero or below.
Over five years, 42.6% of IPOs lost more than half their value from the first close, and the median five-year return was −32.0%.
Do larger IPO companies lose less?
Yes, clearly. Among the 4,112 IPOs with trailing sales of at least $100 million (2025 dollars), only 25.9% fell more than 50% over three years, and the median three-year return from the first close was 2.6%, not −25.7%.
Mia Scarlett: A −25.7% median next to a 21.2% mean is the whole IPO asset class in two numbers. Owning the index of new issues means owning the few outliers that pay for everything else; owning three or four IPOs means you are most likely holding the median.
The Axis IPO Allocation Gap (IAG): Who Actually Earns IPO Returns?
The Axis IPO Allocation Gap (IAG) measures how many percentage points of three-year, market-adjusted return are available only to investors who receive shares at the offer price, compared with investors who buy at the first closing price. A high IAG means the returns in that segment are front-loaded into the allocation. A low IAG means open-market buyers get most of what allocated investors get.
Formula: IAG = (3-year market-adjusted buy-and-hold return from the offer price) − (3-year market-adjusted buy-and-hold return from the first close). Both inputs come from the same IPO sample in the same Ritter table, so the difference isolates the first trading day.
Worked example, all U.S. IPOs 1980–2024: −3.3% (from offer price, Table 16f) − (−20.5%) (from first close, Table 16f) = 17.2 percentage points.
IAG readings by IPO segment
| Segment (U.S. IPOs, 1980–2024) | Market-adjusted from offer | Market-adjusted from first close | IAG (points) | Source |
|---|---|---|---|---|
| All IPOs | −3.3% | −20.5% | 17.2 | Axis calculation on Ritter Table 16f |
| Tech | 15.6% | −12.7% | 28.3 | Axis calculation on Ritter Table 16f |
| Non-tech | −13.9% | −24.9% | 11.0 | Axis calculation on Ritter Table 16f |
| VC-backed, sales over $100M | 41.1% | 11.0% | 30.1 | Axis calculation on Ritter Table 18a |
| Non-VC-backed, sales over $100M | 3.4% | −8.3% | 11.7 | Axis calculation on Ritter Table 18a |
| VC-backed, sales under $100M | −1.2% | −23.5% | 22.3 | Axis calculation on Ritter Table 18a |
| Non-VC-backed, sales under $100M | −33.0% | −45.8% | 12.8 | Axis calculation on Ritter Table 18a |
| Float 10% or less (sales over $100M) | 24.6% | −5.3% | 29.9 | Axis calculation on Ritter Table 18e |
| Float above 40% (sales over $100M) | −3.9% | −12.6% | 8.7 | Axis calculation on Ritter Table 18e |
Axis Intelligence Research finds that the IAG falls steadily as the public float rises: 29.9 points for floats of 10% or less, 21.3 for 10–20%, 16.0 for 20–30%, 11.4 for 30–40% and 8.7 above 40%. Scarcity of tradable shares is where the day-one premium lives.
The best segment for any investor is VC-backed companies with more than $100 million of sales. They beat the market by 11.0 points even from the first close. The allocated investor beat it by 41.1.
The Day-One Capture Ratio (DCR)
A second Axis metric answers a simpler question: how much of the three-year IPO return arrives on the first day? The Day-One Capture Ratio (DCR) divides the mean first-day return by the mean three-year buy-and-hold return from the offer price.
| Segment (U.S. IPOs, 1980–2024) | Mean first-day return | 3-year return from offer | DCR | Source |
|---|---|---|---|---|
| All IPOs | 18.9% | 36.3% | 52.1% | Axis calculation on Ritter Table 16f |
| Tech | 31.2% | 50.1% | 62.3% | Axis calculation on Ritter Table 16f |
| Non-tech | 12.0% | 28.6% | 42.0% | Axis calculation on Ritter Table 16f |
| VC-backed, sales over $100M | 23.4% | 79.4% | 29.5% | Axis calculation on Ritter Table 18a |
| Non-VC-backed, sales over $100M | 10.1% | 43.5% | 23.2% | Axis calculation on Ritter Table 18a |
| VC-backed, sales under $100M | 28.5% | 32.7% | 87.2% | Axis calculation on Ritter Table 18a |
| Non-VC-backed, sales under $100M | 17.5% | 12.8% | 136.7% | Axis calculation on Ritter Table 18a |
A DCR above 100% means the first day paid more than the whole three-year holding period: small non-VC-backed issuers gave back part of their day-one gain. DCR is a ratio of averages, not an average of per-IPO ratios, so it describes each segment rather than a typical individual deal.
Mia Scarlett: The IAG is the number underwriters would rather you didn’t compute. Seventeen points of market-adjusted return accrue to whoever is on the allocation list, and in tech it’s twenty-eight. Anyone pitching IPOs as a retail opportunity should be asked which price the performance statistic assumes.
Which IPOs Perform Best? Sales, Profits, Backers and Valuation
Company size is the most reliable separator in Ritter’s data. IPOs with under $10 million in trailing sales (2025 dollars) posted a −52.3% three-year market-adjusted return; those with $100–499.999 million in sales posted −2.6%, and beat style-matched peers by 5.4%.
IPO returns by pre-IPO sales
| Trailing sales (2025 $) | IPOs | First-day return | 3-year return | Market-adjusted | Style-adjusted | Source |
|---|---|---|---|---|---|---|
| $0–9.999M | 1,739 | 23.6% | −16.2% | −52.3% | −34.0% | Jay R. Ritter, Table 16 |
| $10–19.999M | 647 | 28.0% | 0.0% | −35.8% | −20.6% | Jay R. Ritter, Table 16 |
| $20–49.999M | 1,362 | 24.7% | 12.5% | −30.1% | −9.6% | Jay R. Ritter, Table 16 |
| $50–99.999M | 1,395 | 18.9% | 27.6% | −15.3% | −4.6% | Jay R. Ritter, Table 16 |
| $100–499.999M | 2,610 | 15.5% | 39.1% | −2.6% | 5.4% | Jay R. Ritter, Table 16 |
| $500M and up | 1,500 | 10.3% | 31.8% | −4.4% | −3.0% | Jay R. Ritter, Table 16 |
The smallest companies deliver the largest first-day returns and the worst long-run returns. The pattern repeats in every cut of the data.
Profitable vs unprofitable IPO returns
Unprofitable issuers averaged a 26.5% first-day return and −30.7% three years later on a market-adjusted basis. Profitable issuers popped 13.3% and trailed the market by 13.0%. Profitable companies with sales above $100 million beat style-matched peers by 5.4%; unprofitable ones below $100 million trailed them by 29.6%.
VC-backed vs buyout-backed vs unsponsored IPOs
| Sponsor type | IPOs | First-day return | 3-year return | Market-adjusted | Style-adjusted | Source |
|---|---|---|---|---|---|---|
| VC-backed | 3,261 | 28.8% | 18.5% | −15.4% | −4.3% | Jay R. Ritter, Table 17a |
| Growth capital-backed | 472 | 14.6% | 42.5% | −0.9% | 4.9% | Jay R. Ritter, Table 17a |
| Buyout-backed | 1,250 | 9.5% | 29.4% | −1.4% | −0.7% | Jay R. Ritter, Table 17a |
| No financial sponsor | 4,270 | 14.6% | 14.0% | −32.2% | −16.3% | Jay R. Ritter, Table 17a |
Sponsor-backed IPOs are the better long-run bet, and buyout-backed deals come closest to matching the market, with the smallest first-day pop. IPOs without a financial sponsor trailed the market by 32.2 points.
IPO valuation and long-run returns
Among IPOs with sales above $100 million, those priced at more than 40 times sales at the first close returned a mean 93.6% on day one and −44.8% over three years, or −58.5% against the market. Those under 5 times sales popped 8.9% and returned 41.3%, just 1.3 points behind the market.
| Price-to-sales at first close (sales over $100M) | IPOs | First-day return | 3-year return | Market-adjusted | Source |
|---|---|---|---|---|---|
| Under 5x | 3,384 | 8.9% | 41.3% | −1.3% | Jay R. Ritter, Table 18d |
| 5x–10x | 395 | 24.1% | 20.7% | −8.0% | Jay R. Ritter, Table 18d |
| 10x–20x | 198 | 36.6% | 18.5% | −7.7% | Jay R. Ritter, Table 18d |
| 20x–40x | 87 | 55.9% | 2.9% | −16.3% | Jay R. Ritter, Table 18d |
| Above 40x | 46 | 93.6% | −44.8% | −58.5% | Jay R. Ritter, Table 18d |
Axis Intelligence Research does not compute an IAG for valuation buckets. Ritter assigns IPOs to price-to-sales groups using the first-close price in one panel and the offer price in the other, so the two panels hold different companies, and subtracting them would compare unlike samples. Our tech IPO valuation analysis, including the Tech IPO Froth Index, sits on the tech IPO statistics page.
Dual-class IPO returns
Dual-class IPOs, which give founders super-voting shares, trailed the market by 6.7 points over three years; single-class IPOs trailed it by 22.0. Among tech companies, the 309 dual-class IPOs beat the market by 13.8 points.
Mia Scarlett: Every one of these tables says the same thing in a different accent: size, profits, a sponsor with its own reputation on the line, and a sane multiple. The governance critics of dual-class shares may be right about accountability. The return data simply hasn’t punished it.
Do First-Day IPO Pops Predict Long-Term Returns?
A falling first day does. Ritter’s 654 “broken” U.S. IPOs from 2001 to 2024, which closed below their offer price, returned −32.0% against the market over three years, and 442 of them (67.6%) posted negative three-year returns. IPOs with a positive first day trailed the market by 8.5 points.
| First-day outcome (2001–2024) | IPOs | Mean first-day return | 3-year return | Market-adjusted | Style-adjusted | Source |
|---|---|---|---|---|---|---|
| Negative (broken) | 654 | −8.9% | −3.8% | −32.0% | −23.2% | Jay R. Ritter, Table 19e |
| Exactly zero | 153 | 0.0% | 10.4% | −15.8% | −3.8% | Jay R. Ritter, Table 19e |
| Positive | 1,927 | 29.6% | 17.5% | −8.5% | −10.3% | Jay R. Ritter, Table 19e |
The share of broken deals swings with the market. In 2023, 53.7% of U.S. IPOs closed their first day below the offer price; in 2025 the figure was 24.4%, alongside a 29.3% mean first-day return tracked on our IPO statistics 2026 hub, per Ritter’s Initial Public Offerings: Updated Statistics.
Pricing above the range and the first-day pop
From 2020 to 2025, 29% of U.S. IPOs priced above their initial file range, 55% within it and 16% below it. Deals priced above the range gained a mean 51% on day one, against 21% within the range and 6% below. Since 1980 the above-range figure is 50%.
Penny-stock IPOs: the biggest pops and the worst returns
IPOs priced below $5 from 2001 to 2024 averaged a 67.1% first-day return and −82.6% against the market over three years. Deals priced at exactly $5 trailed the market by 91.3 points. IPOs priced above $5 trailed it by 11.8.
Mia Scarlett: A broken IPO is the market’s first honest audit of the roadshow, and it is right about two times in three. An above-range pricing tells you demand was deep; it doesn’t tell you the price was right, which is why the 51% day-one gain and the long-run shortfall can sit in the same sample.
Does the Underwriter Matter for IPO Returns?
It matters most at the bottom of the league table. IPOs led by lower-tier underwriters from 2012 to 2024 posted a 37.1% first-day return, the highest of any group, and −89.1% against the market over three years. Morgan Stanley-led IPOs trailed the market by 0.5 points, a gap Axis Intelligence Research calculates at 88.6 points.
IPO returns by lead-left underwriter, 2012–2024
| Lead-left underwriter | IPOs | First-day return | 3-year return | Market-adjusted | Style-adjusted | Source |
|---|---|---|---|---|---|---|
| Goldman Sachs | 299 | 26.1% | 9.5% | −23.7% | −19.6% | Jay R. Ritter, Table 11a |
| JP Morgan | 248 | 24.7% | 18.5% | −13.3% | −3.0% | Jay R. Ritter, Table 11a |
| Morgan Stanley | 232 | 26.5% | 30.8% | −0.5% | 3.9% | Jay R. Ritter, Table 11a |
| BOA-Merrill | 140 | 23.8% | 18.6% | −16.1% | −1.1% | Jay R. Ritter, Table 11a |
| Jefferies | 99 | 23.8% | 0.6% | −33.1% | −25.7% | Jay R. Ritter, Table 11a |
| UBS/Credit Suisse | 89 | 10.9% | 16.9% | −17.4% | −19.6% | Jay R. Ritter, Table 11a |
| Citigroup | 83 | 8.0% | 24.6% | −8.6% | −29.8% | Jay R. Ritter, Table 11a |
| Barclays | 53 | 15.4% | 30.0% | −0.4% | 8.0% | Jay R. Ritter, Table 11a |
| Lower-tier underwriters | 210 | 37.1% | −54.5% | −89.1% | −60.4% | Jay R. Ritter, Table 11a |
| All | 1,645 | 23.4% | 7.6% | −25.5% | −16.7% | Jay R. Ritter, Table 11a |
Ritter’s market benchmark here is the Morningstar US Total Market Index. The bank table reflects the deals each bank led, which depends on the clients it wins, so it measures deal mix as much as pricing skill.
Mia Scarlett: Read the bank table as a credit analyst reads a loan book. A 37% average pop followed by an 89-point market shortfall is not an underwriter mispricing a few deals; it is a deal mix. The names on the prospectus cover page are a data point worth reading before the numbers inside.
IPO vs Direct Listing vs SPAC: Which Route Has Returned More?
The traditional IPO sits in the middle. From 1999 to 2021, 3,404 bookbuilt U.S. IPOs returned −3.4% over three years, or −17.5% market-adjusted. Ritter’s 23 auction IPOs from 1999 to 2008 returned 87.5% and beat the market by 79.2 points, and six of those 23 posted three-year returns of at least 150%.
| Listing route | Listings | First-day return | 3-year return | 3-year market-adjusted | 5-year return | 5-year market-adjusted | Source |
|---|---|---|---|---|---|---|---|
| Auction IPOs (1999–2008) | 23 | 12.7% | 87.5% | 79.2% | 89.2% | 71.5% | Jay R. Ritter, Table 13b |
| Direct listings (2018–2021) | 12 | 2.1% | −3.6% | −36.4% | 131.9% | 65.6% | Jay R. Ritter, Table 13b |
| Other IPOs (1999–2021) | 3,404 | 30.2% | −3.4% | −17.5% | 11.8% | −20.3% | Jay R. Ritter, Table 13b |
Direct-listing first-day returns are measured from open to close, because no shares trade at a reference price.
De-SPAC returns
The 451 SPAC mergers completed from 2012 to 2022 returned a mean −46.3% in their first year, −49.4% against the market, and −57.7% over three years, a 74.7-point market shortfall. The 43 de-SPACs of 2025 returned −55.7% through year-end, measured from an assumed $10 price.
Direct listings in 2026
Ritter’s table logs 18 U.S. direct listings dated between January 27 and August 25, 2026. Axis Intelligence Research counts 11 of the 18 closing their first day below the opening price. Ritter notes that direct listings since 2022 have generally been microcap stocks, unlike the Spotify, Palantir and Roblox listings of 2018–2021.
Mia Scarlett: EY reports that SPAC formation and dual-track IPO/SPAC processes are back in 2026. The de-SPAC ledger is the context that belongs next to that news: a 74.7-point three-year shortfall is not a cycle, it is a structural cost of the product.
How Many IPOs Are Withdrawn or Postponed?
HKEX publishes a monthly count of listing applications that fail to list. From January to July 2026, HKEX processed 838 applications; 131 listed, 12 were approved and awaiting listing, 485 were still under review, and 210 lapsed, were rejected, returned or withdrawn, according to the HKEX July 2026 report on IPO applications.
Hong Kong IPO application outcomes, 2024–2026
| Period | Applications processed | New applications | Listed | Approved, pending | Under processing | Lapsed, rejected, returned or withdrawn | Source |
|---|---|---|---|---|---|---|---|
| 2024 | 250 | 171 | 107 | 20 | 64 | 59 | HKEX, Dec. 2024 report |
| 2025 | 612 | 516 | 168 | 24 | 321 | 99 | HKEX, Dec. 2025 report |
| 2026 (Jan–Jul) | 838 | 459 | 131 | 12 | 485 | 210 | HKEX, July 2026 report |
Axis Hong Kong IPO drop-out metrics
| Metric | 2024 | 2025 | 2026 (Jan–Jul) | Formula | Source |
|---|---|---|---|---|---|
| Drop-out share of resolved applications | 35.5% | 37.1% | 61.6% | dropped ÷ (dropped + listed) | Axis calculation on HKEX |
| Attrition rate | 23.6% | 16.2% | 25.1% | dropped ÷ processed | Axis calculation on HKEX |
| Listing conversion rate | 42.8% | 27.5% | 15.6% | listed ÷ processed | Axis calculation on HKEX |
Axis Intelligence Research finds that more Hong Kong applications dropped out (210) than listed (131) in the first seven months of 2026. Two mechanics explain much of it. An HKEX application lapses automatically after six months, and the exchange processed a surge of filings in 2025: 516 new applications against 171 in 2024. Rejections play almost no part: HKEX reports zero rejections and one returned application for 2026 year to date.
Processing time has lengthened too. The median application heard in the 12 months to July 2026 took 106 business days to reach its hearing-bundle letter, against 92 in the 12 months to December 2025. HKEX says a surge in filings pushed some reviews beyond its enhanced timeframe.
U.S. IPO withdrawals
The SEC does not publish an aggregate count of withdrawn IPOs. Each withdrawal is an individual Form RW filed under Securities Act Rule 477, typically citing market conditions and confirming that no securities were sold, as in this Form RW filed on EDGAR. Axis Intelligence Research does not republish third-party U.S. withdrawal tallies, because no primary source publishes one and private tallies apply different inclusion rules for SPACs and penny stocks.
Mia Scarlett: A lapse is not a failure in the way a pulled roadshow is: an applicant that refiles within three months counts as a renewal under HKEX’s definitions. But 210 drop-outs against 131 listings is still the truest measure of how selective the 2026 window is, and it’s the number that never appears in the proceeds headlines.
What Past IPO Performance Means for the 2026 IPO Class
The 2026 class is dominated by larger deals. EY’s Q2 2026 Global IPO Trends reports that 12 U.S. IPOs raised more than $1 billion each in the first half of 2026, up from four a year earlier, and that recent deals have generally priced and traded well.
Ritter’s data suggests where to look for trouble: small floats, high multiples, unprofitable small issuers and lower-tier bookrunners. It also suggests where the 2026 class looks healthier: larger sales bases and sponsor backing, the two groups with the best long-run records.
EY also publishes average aftermarket returns for the 2025 and 1H 2026 classes by region. Axis Intelligence Research has not included them: the chart labels in EY’s report cannot be matched to their data series with certainty, and we do not publish a figure whose meaning we cannot verify. Country-level listing counts are covered in our IPO statistics by country, and every priced AI listing is logged in the AI IPO Tracker.
Mia Scarlett: The 2026 class will get its three-year report card in 2029. Until then, the only honest forecast is the base rate: most individual IPOs trail the market, the average is rescued by a few names, and the allocation list decides who gets the first seventeen points.
Methodology
Collection. Axis Intelligence Research assembled this dataset on September 16, 2026, from seven primary documents, each fetched and read: Jay R. Ritter’s Initial Public Offerings: Updated Long-run Statistics (University of Florida, August 27, 2026) and Initial Public Offerings: Updated Statistics (September 14, 2026); EY Q2 2026 Global IPO Trends (July 2026, Dealogic data); HKEX’s Report on Initial Public Offering Applications, Delisting and Suspensions for December 2024, December 2025 and July 2026; and a Form RW on SEC EDGAR illustrating the U.S. withdrawal mechanism. Every figure on this page is a row in the downloadable CSV with source URL and retrieval date.
Sample definitions. Ritter’s U.S. sample covers operating-company IPOs with an offer price of at least $5, excluding ADRs, unit offers, SPACs, REITs, closed-end funds, banks and S&Ls, natural resource partnerships, small best-efforts deals and stocks not listed on CRSP within six months. Buy-and-hold returns include dividends and run from the first closing price (or the offer price, where stated) to the earlier of the three-year anniversary, delisting, or December 31, 2025. Market-adjusted returns subtract the compounded CRSP value-weighted index (Morningstar US Total Market Index in Table 11a); style-adjusted returns subtract a matched non-issuer of similar size and book-to-market ratio. HKEX counts include Main Board, GEM and investment-vehicle applications; “dropped” combines lapsed, rejected, returned and withdrawn applications, which HKEX reports as one line.
Axis formulas.
- IPO Allocation Gap (IAG) = 3-year market-adjusted return from offer price − 3-year market-adjusted return from first close, taken from the same table and sample (Tables 16f, 18a, 18e).
- Day-One Capture Ratio (DCR) = mean first-day return ÷ mean 3-year buy-and-hold return from offer price × 100 (Tables 16f, 18a).
- Share of IPOs with zero or negative 3-year returns = sum of Table 16e counts in the two lowest buckets ÷ 9,195.
- 2020–2023 cohort average = simple mean of four cohort market-adjusted returns (not weighted by deal count).
- HKEX drop-out share = dropped ÷ (dropped + listed); attrition = dropped ÷ processed; conversion = listed ÷ processed.
- 2026 direct-listing counts = Axis count of 2026-dated rows in Ritter Table 13a. All calculations were run twice, in floating point and in exact rational arithmetic.
Scope. Long-run returns cover U.S. listings only; no exchange outside Hong Kong publishes application drop-outs on a comparable basis. The 2026 Hong Kong figures cover seven months and are compared with full years, so the 2026 attrition and conversion rates will move as pending applications resolve. Returns for the 2023 and 2024 classes cover less than three years. IAG is not computed where Ritter’s panels hold different samples.
About This Dataset
File: ipo-performance-statistics.csv, 516 rows in long format (metric, value, unit, as_of_date, geography, segment, period) with full provenance columns (source_org, source_document, source_url, retrieved_date, is_primary, axis_calculated, method_note, license).
Coverage: U.S. IPO first-day, one-year, three-year and five-year returns by era, class year, sales, profitability, sponsor type, valuation, float, share structure, first-day outcome, offer price, file-range pricing and lead underwriter; direct listing, auction IPO and de-SPAC returns; Hong Kong IPO application outcomes for 2024, 2025 and January–July 2026; and every IAG and DCR reading.
License: CC BY 4.0. Reuse freely with attribution: Axis Intelligence Research, IPO Performance Statistics 2026, 2026.
Also available on: Hugging Face · Kaggle · GitHub (Axis Intelligence Research).
How to Cite This Page
APA Axis Intelligence Research, & Scarlett, M. (2026, September 16). IPO performance statistics 2026: First-day pops, three-year returns and withdrawn IPOs. Axis Intelligence. https://axis-intelligence.com/ipo-performance-statistics/
MLA Axis Intelligence Research, and Mia Scarlett. “IPO Performance Statistics 2026: First-Day Pops, Three-Year Returns and Withdrawn IPOs.” Axis Intelligence, 16 Sept. 2026, axis-intelligence.com/ipo-performance-statistics/.
Chicago Axis Intelligence Research, and Mia Scarlett. “IPO Performance Statistics 2026: First-Day Pops, Three-Year Returns and Withdrawn IPOs.” Axis Intelligence, September 16, 2026. https://axis-intelligence.com/ipo-performance-statistics/.
IPO Performance Questions Investors Are Asking
Is it better to buy an IPO at the offer price or on the first day of trading?
The offer price, by a wide margin. U.S. IPOs from 1980 to 2024 trailed the market by 3.3 points over three years from the offer price and by 20.5 points from the first close, per Jay R. Ritter. Axis Intelligence Research measures that difference as a 17.2-point IPO Allocation Gap (IAG).
What percentage of IPOs lose money?
About 60%. Axis Intelligence Research calculates that 60.0% of U.S. IPOs from 1975 to 2021 had a zero or negative three-year return from their first closing price, and 38.5% fell more than 50%, based on Jay R. Ritter’s Table 16e.
When do IPO stocks perform worst after listing?
In the second half of the first year and the second year. U.S. IPOs from 1980 to 2024 returned 4.9% in their first six months and 0.0% in the next six, then trailed size-matched firms by 7.9 points in year two, before pulling ahead in year four, per Jay R. Ritter.
Does a big first-day IPO pop mean the stock will keep rising?
No. Pops and long-run returns tend to move in opposite directions. IPOs valued above 40 times sales at the first close gained 93.6% on day one and trailed the market by 58.5 points over three years, per Ritter’s data on companies with sales above $100 million.
What happens to IPOs that fall on their first day?
They usually keep underperforming. Of 654 U.S. IPOs from 2001 to 2024 that closed their first day below the offer price, 442 (67.6%) had negative three-year returns, and the group trailed the market by 32.0 points, according to Jay R. Ritter.
Why do so many Hong Kong IPO applications lapse or get withdrawn in 2026?
Mostly because of volume and the six-month lapse rule. HKEX processed 838 applications from January to July 2026; 210 lapsed, were rejected, returned or withdrawn, and 131 listed. HKEX recorded zero rejections, and an applicant refiling within three months of a lapse counts as a renewal.
Do SPAC mergers perform better than traditional IPOs?
No. De-SPAC companies from 2012 to 2022 returned −57.7% over three years, a 74.7-point shortfall to the market, per Ritter’s Table 15c. Traditional U.S. IPOs from 1999 to 2021 trailed the market by 17.5 points over three years.
Which type of IPO has the best long-term track record?
Larger, sponsor-backed companies. VC-backed IPOs with more than $100 million in trailing sales beat the market by 11.0 points over three years from the first close, and growth capital-backed IPOs came within 0.9 points of it, according to Jay R. Ritter.
Related research: IPO Statistics 2026 · IPO Statistics by Country · Tech IPO Statistics · AI IPO Tracker