Special Report · IPO Investing

Investing in IPOs:
What 45 Years of Data Says

Across 9,343 U.S. IPOs (1980–2025), the typical newly public company lags the market over the following three years. The losses are not random: they cluster in the most expensive, least profitable, hype-driven deals.

Source: Jay R. Ritter, University of Florida, “Initial Public Offerings: Updated Statistics” (May 2026) · Returns through Dec 31, 2025
Start with the base rate

01The base rate: IPOs underperform the market

Buying an IPO at the first-day closing price and holding for three years (1980–2024, 9,253 deals) returned +19.1% on average, but the market returned far more over the same windows.

−20.5%
3-yr return vs. market (avg)
−25.7%
Median 3-yr return
38.5%
IPOs that lose >50% in 3 yrs
~16%
IPOs that more than double
The median is the story, not the average
A handful of huge winners pull the average up, but the median IPO loses about a quarter of its value in three years. Most IPOs disappoint; a few make the headlines. (Returns from the first-day close, what a retail buyer actually pays.)
The single most useful chart

02The richer the IPO, the worse the 3-year return

Among large IPOs (trailing revenue above $100M), sorting by price-to-sales ratio at the first-day close reveals a clean, monotonic penalty for paying up. The most expensive bucket is a wealth-destruction machine.

3-year return vs. market, by price-to-sales ratio
-66-47-29-117-1.3P/S <5-85–10-7.710–20-16.320–40-58.5>40
Large IPOs ($100M+ revenue), 1980–2024. Bars below the zero line lag the CRSP value-weighted market over 3 years. Measured from first close.
IPOs priced above 40× revenue
−58.5%
underperformance vs. the market over 3 years (style-adjusted: −75.9%).
+93.6%
Avg first-day pop, >40× group
6 of 7
Beaten by the market after
The headline takeaway
IPOs that come public at a valuation above 40× revenue jump nearly +94% on day one, then underperform the market by ~58% over the next three years. The day-one excitement is precisely what you are paying for, and it is a bad trade for the buyer.
What separates winners from losers

03Profits and scale matter more than the story

The IPO penalty is concentrated in unprofitable, tiny-revenue companies. Profitable, larger businesses roughly keep pace with the market.

Profitability, 3-yr vs. market
-35-25-14-46-30.7-23.4Unprofitable-13+1.8Profitable
Market-adjustedStyle-adjusted (vs. matched peer)
Profitable IPOs roughly match a size/value-matched peer (style-adjusted +1.8%). Unprofitable ones lag badly.
Pre-IPO revenue, 3-yr vs. market
-59-42-26-106-52.3<$10M-30.1$10–50M-15.3$50–100M-2.6$100–500M-4.4>$500M
Sales in 2025 dollars. The smallest-revenue cohort is the worst; bigger pre-IPO revenue means a smaller penalty.
−34.3%
Revenue <$100M (vs. market)
−3.2%
Revenue >$100M (vs. market)
+1.8%
Profitable, style-adjusted
The takeaway
Fundamentals beat narrative. A profitable business with $100M+ in revenue roughly keeps pace with the market; the deep losses are concentrated in tiny, money-losing “story” IPOs. Insist on real profits and real scale.
Who really captures the pop

04The first-day pop is a seller’s number, not a buyer’s

The famous IPO “pop” is money transferred from the issuing company to whoever got an allocation at the offer price. If you buy at the open, you have already missed it, and you are now exposed to the long-run underperformance.

+19.0%
Avg first-day return, 1980–2025
$250B
Money left on the table (cumulative)
+29.3%
Avg first-day pop in 2025
+125%
Avg pop for the hottest IPOs
Offer price vs. first close
Measured from the offer price, the 3-year market-adjusted return is roughly flat to slightly negative (−3.3%). Measured from the first close, the realistic retail entry, it drops to −20.5%. That gap is the first-day pop.
Signals hiding in the deal terms

05The price tag and the sponsor are signals too

Two things you can read straight off the prospectus carry real information: the offer price per share, and who is backing the company. Cheap shares and no institutional sponsor are among the worst combinations in the data.

3-year return vs. market, by offer price
−11.8%
Above $5
−91.3%
Exactly $5
−82.6%
Below $5 (penny)
2,913 IPOs, 2001–2024, from first close. Low-priced and penny-stock IPOs are structural wealth-destroyers; a sub-$8 file-price midpoint averaged −86%.
3-year return vs. market, by sponsor type
-36-26-16-64-0.9Growth cap.-1.4Buyout (PE)-15.4VC-32.2No sponsor
9,253 IPOs, 1980–2024, from first close. Any serious financial sponsor helped; deals with no sponsor at all lagged the market by ~32%.
The takeaway
The prospectus leaks signals before you ever value the company. Low-priced and penny (sub-$5) offerings and deals with no institutional sponsor are among the worst buckets in the data; a serious buyout or growth-capital backer is a quiet vote of confidence.
When the crowd is most excited

06The hottest IPO years produced the worst cohorts

3-year return vs. market, by IPO cohort year
-99-72-44-1711-32.51999-30.92000-25.12019-78.62020-68.62021-72.22022-88.32023-33.22024-20.51980–24
Hot-year cohort (vs. market)All IPOs avg, 1980–2024 (vs. market)
The dot-com peak (1999–2000) and the 2020–2023 boom-and-bust both produced cohorts that trailed the market by 30–88% over three years. Yellow bar: full-sample average (−20.5%), all IPO cohorts 1980–2024.
SPACs & deSPACs are even worse
−57.7%
deSPAC 3-yr return (avg)
−74.7%
deSPAC 3-yr vs. market
Across 451 completed SPAC mergers (2012–2022), buying the combined company at its first close and holding three years lost more than half your money, and trailed the market by ~75%.
Distribution of 3-year outcomes
3-year buy-and-hold returnShare of IPOs
Lost more than 50%38.5%
Down 0% to 50%21.5%
Up 0% to 100%24.0%
More than doubled16.0%
9,195 IPOs, 1975–2021, from the first close. Six in ten IPOs are underwater after three years.
The takeaway
Buy when the crowd is bored, not euphoric. The hottest IPO windows (1999–2000, 2020–2023) and SPACs / deSPACs produced the worst cohorts of all. If everyone is excited about IPOs, that is the signal to be more demanding, not less.
The exceptions that worked

07It is not all bad: where IPOs have beaten the market

Profitable tech, real revenue
+44%
3-yr vs. market (offer, ex-bubble)
Larger, profitable technology IPOs (excluding 1999–2000) have been the bright spot.
Low public float
+24.6%
Float ≤10%, from offer price
Tightly-floated large IPOs outperformed; high-float deals (>40%) lagged.
Backed & scaled
+11%
VC-backed, >$100M revenue
VC backing helps only when paired with real revenue scale.
Founder-controlled tech
+13.8%
Dual-class tech, 3-yr vs. market
Dual-class tech beat the market; dual-class deals overall lagged just −7% vs. −22% for single-class.
The takeaway
IPOs can win, but only a specific kind. The bright spots share a profile: profitable tech with real revenue, a tight float, and committed founders or sponsors. The edge comes from the fundamentals, not from the IPO label itself.
Putting it into practice

08Investor checklist

Characteristics that improved oddsTilt toward
  • +Profitable companies (or a clear, near-term path to profit)
  • +Trailing revenue above $100M, scale lowers the penalty
  • +Price-to-sales below ~5–10× at the price you actually pay
  • +Buying near the offer price, not chasing the first-day pop
  • +Larger, established tech with genuine revenue
  • +A serious sponsor (buyout / growth-capital) behind the deal
  • +Founder-controlled (dual-class) tech with real fundamentals
Red flags that predicted underperformanceBe wary of
  • ×Valuations above 40× revenue, the worst long-run bucket (−58%)
  • ×Unprofitable, sub-$100M revenue “story” stocks
  • ×Buying into the first-day spike on the hottest deals
  • ×SPACs / deSPACs, structurally poor 3-year returns
  • ×Low-priced and penny (sub-$5) offerings, the worst bucket of all
  • ×Companies with no institutional sponsor behind them
  • ×Piling in during euphoric IPO windows (1999, 2020–21)
Bottom line
IPOs are not a free lottery ticket. The market pays the day-one pop to the seller; the buyer inherits a stock that, on average, lags for years. Skew toward profitable, real-revenue, reasonably-valued businesses, and treat anything priced above 40× revenue with deep skepticism.
The biggest deals, by price-to-sales

09What the marquee tech IPOs actually cost

The marquee technology IPOs priced across an enormous valuation range, from about 2× sales (Nvidia) to nearly 95× (SpaceX). The pattern is striking: the greatest compounders ever — Nvidia, Microsoft, Apple — came public cheap and profitable, while SpaceX’s June 2026 listing at a ~$1.77T valuation is the most expensive of them all.

Price-to-sales at IPO — marquee tech deals (incl. SpaceX, 2026)
0×27×53×80×106×2×Nvidia19994×Microsoft19867×Uber201914×Apple198015×Tesla201016×Google200420×Alibaba201428×Facebook201228×Amazon199728×Palantir2020 DL33×Robinhood202145×Twitter201359×Snap201767×Coinbase2021 DL95×SpaceX2026 IPO
Reasonable (<10× sales)Rich (10–40×)Danger zone (>40×)SpaceX (June 2026 IPO)
Price-to-sales at the IPO valuation (offer-price market cap ÷ trailing fiscal-year total revenue), from each company’s SEC prospectus / pricing release. The multiple is methodology-sensitive: e.g. Google is ~16× on 2003 gross revenue ($1.47B) but ~24× on net/ex-TAC revenue, and Microsoft is ~4× on trailing revenue (~$172.5M) but ~5.5× on FY1985. “DL” marks direct listings (Palantir, Coinbase), at their debut-day fully-diluted cap. Legacy marks: Apple ~$1.7B ÷ $117.9M (FY1980) → ~14×; Nvidia ~$343M ÷ $158M (FY1999) → ~2×. SpaceX uses its June 2026 S-1: a ~$1.77T valuation ($135/share, ~$75B raised) against FY2025 revenue of $18.7B → ~95×.
2–14×
Nvidia, Microsoft & Apple at their IPOs
~95×
SpaceX — P/S at its ~$1.77T IPO
4 of 15
Marquee tech IPOs priced above 40× sales
The Bastion read
The three best-performing tech stocks of all time — Nvidia (~2×), Microsoft (~4×) and Apple (~14×) — were among the cheapest deals on this chart at IPO, and all three were profitable. SpaceX is asking ~95× sales on $18.7B of revenue — far beyond the 40× line where IPOs historically lag the market by ~58% over three years (section 02). The lesson of this report in one picture: the giant compounders came public cheap; the richly-priced debuts (Twitter, Snap, Coinbase) are the cohort the data warns about.
Score it before you buy

The Bastion IPO Health Score

Fourteen questions you can answer before the deal prices. Get a single 0–100 read on how healthy the IPO looks.

Open the IPO Health Score
Source: Jay R. Ritter, University of Florida, “Initial Public Offerings: Updated Statistics” (May 2026). All long-run figures are 3-year buy-and-hold returns, measured from the first-day closing price unless noted, versus the CRSP value-weighted index. “Style-adjusted” compares each IPO to a size- and book-to-market-matched peer.