94 US IPOs · 2020–2025 · Data as of May 2026

IPO Valuation
Intelligence

The valuation dead zone, power-law winners, and what six years of IPO data reveals about when — and at what price — to go public.

0
Total IPOs
0%
Win Rate
-0%
Median Return
+0%
Best Return (CRDO)
Return distribution across all 94 IPOs
← All lossesBreak-evenAll gains →
Core Thesis

The Valuation Ladder

Every IPO bucket from sub-$1B to $40B+, ranked by win rate and median return. The dead zone at $10–20B is not an anomaly — it's a structural trap.

<$1B6 companies
33%
Win Rate
-3%
Median
33%
Big Wins (>100%)
17%
Big Loss (>50%↓)

GPCR +962%APGE +442%TWG -3%TXO -3%LXEO -29%CRGX -79%
InsightThe $1–2B bucket has the best risk-adjusted entry: 60% win rate, +109% median. The $10–20B bucket is the structural floor: 22% win rate, -41% median, zero doubles. Not bad luck — structural mispricing at the premium-but-not-category-defining valuation.
Power Law Analysis

The Distribution Is Not Normal

IPO Valuation ($B) vs Total Return — notable IPOs labeled
InsightNo correlation between IPO valuation and return — but there's a band of destruction at $10–20B (all red dots). The two superstars (CRDO at $1.6B, PLTR at $21.8B) float far above the field — separated from the pack by a factor of 10+.
Best & Worst

Hall of Fame vs Hall of Shame

The top 10 and bottom 10 companies by total return since IPO. Bar width is log-scaled for winners; linear for losers.

Top 10 Winners
1
CRDO
+2,162%
2022
2
PLTR
+1,415%
2020
3
GPCR
+962%
2023
4
ATAT
+948%
2022
5
HOOD
+781%
2021
6
ALAB
+558%
2024
7
NXT
+454%
2023
8
APGE
+442%
2023
9
CRCL
+422%
2025
10
KGS
+405%
2023
Bottom 10 Losers
1
DIDI
-99%
2021
2
BIRD
-99%
2021
3
VRM
-99%
2020
4
WISH
-99%
2020
5
AMWL
-97%
2020
6
GDRX
-95%
2020
7
BMBL
-94%
2021
8
GEMI
-89%
2025
9
STUB
-88%
2025
10
ROOT
-87%
2020
InsightThe top 10 winners span semiconductors, data infra, and fintech — and include both tiny ($1.6B CRDO) and mega ($21.8B PLTR) IPOs. The bottom 10 are almost entirely 2021 marketplace and consumer tech names that priced at peak ZIRP multiples. Vintage and sector compounded the destruction.
Vintage Analysis

The Year You IPO'd Determined Your Fate

IPO count (bars) + win rate (line) by vintage year
Insight2021 had 27 IPOs (most of any year) but only 19% win rate — the worst vintage in the dataset. 2022 had just 5 IPOs and 60% win rate with +88% median return. Volume and quality move in opposite directions.
Cross-Tab Matrix

Year × Valuation Bucket Heatmap

Win rate per year-bucket cell. Dark red = mostly losses. Bright green = mostly wins. Empty = no IPOs that year in that bucket. Hover for detail.

Year<$1B$1-2B$2-3B$3-5B$5-7B$7-10B$10-20B$20-40B$40B+
2020
—
75%
n=4
33%
n=3
33%
n=3
0%
n=4
0%
n=2
25%
n=4
100%
n=3
100%
n=1
2021
—
0%
n=1
—
40%
n=5
0%
n=5
20%
n=5
20%
n=5
50%
n=2
0%
n=4
2022
0%
n=1
100%
n=2
—
—
0%
n=1
100%
n=1
—
—
—
2023
50%
n=4
100%
n=1
67%
n=3
100%
n=1
—
33%
n=3
—
—
100%
n=1
2024
0%
n=1
—
0%
n=1
100%
n=1
100%
n=4
50%
n=2
—
—
0%
n=1
2025
—
0%
n=2
0%
n=1
25%
n=4
50%
n=2
25%
n=4
—
50%
n=2
—
Insight2021's $5–10B range is almost entirely red — the most populated IPO band in the worst vintage. 2022's $1–2B cell shows CRDO's 100% win rate. The $10–20B column is red across every year that has data — the dead zone transcends vintage cycles.
All 94 Companies

The Full Mosaic

Every IPO sorted by return, best to worst. Green = winner, red = loser. Filter by vintage year.

CRDO
+2162%
PLTR
+1415%
GPCR
+962%
ATAT
+948%
HOOD
+781%
ALAB
+558%
NXT
+454%
APGE
+442%
CRCL
+422%
KGS
+405%
RDDT
+393%
CAVA
+276%
ACMR
+230%
STEP
+217%
ARM
+196%
CRWV
+174%
LMND
+164%
RBRK
+121%
DASH
+120%
FROG
+119%
ONON
+100%
TPG
+88%
LI
+84%
ABNB
+78%
SNOW
+59%
UPST
+54%
TEM
+46%
GLBE
+38%
DUOL
+36%
VOYG
+32%
FIG
+30%
HSAI
+17%
SRVT
+15%
WAY
+13%
CART
+3%
RBLX
+0%
LINE
-1%
DNB
-2%
TWG
-3%
TXO
-3%
CHYM
-3%
MDLN
-4%
XPEV
-7%
JAMF
-7%
BLCO
-9%
TOST
-10%
COMP
-10%
U
-10%
BIRK
-14%
COIN
-15%
MNTN
-20%
HNGE
-20%
SLDE
-20%
ETOR
-25%
KLAR
-27%
RYAN
-28%
LXEO
-29%
CPNG
-30%
NCNO
-32%
SUMO
-37%
WRBY
-38%
DOCS
-41%
NRDS
-47%
MNDY
-48%
DNUT
-48%
KVYO
-49%
ZI
-56%
VIA
-60%
FIGR
-60%
BRZE
-61%
IBTA
-68%
VG
-69%
ODD
-70%
ASAN
-72%
RIVN
-75%
COUR
-77%
AI
-78%
CRGX
-79%
MQ
-79%
UDMY
-81%
LZ
-85%
SG
-85%
TASK
-85%
BIGC
-85%
ROOT
-87%
STUB
-88%
GEMI
-89%
BMBL
-94%
GDRX
-95%
AMWL
-97%
WISH
-99%
VRM
-99%
BIRD
-99%
DIDI
-99%
>+200%
+100 to 200%
0 to +100%
-30 to 0%
-30 to -60%
<-60%
Sector Analysis

Where You Played Mattered as Much as When

Sector median return — sectors with 3+ IPOs
Sector win rate vs median return (bubble size = n)
InsightSemiconductors (top-right quadrant: 100% win rate, +394% median) and AI/Data Infra stand alone. SaaS/Software — despite being the largest category (15 IPOs) — is stuck in the bottom-left with -10% median. The AI tailwind created a structural sector divergence.
Capital Impact

Dollars Created and Destroyed by Bucket

Net market cap change since IPO, aggregated by valuation bucket ($B). Green bars = value created. Red bars = value destroyed. Gold line = net.

InsightThe $20–40B bucket shows the largest gross creation (PLTR compounded from $21.8B to $330B+). The $40B+ bucket shows massive destruction — RIVN and DiDi alone vaporized over $100B in investor capital. The $1–2B sweet spot quietly generates strong net value relative to capital raised.
Sensitivity Analysis

How Dependent Is the Story on a Few Outliers?

The mean return is almost entirely driven by a handful of outliers. The median is robust. Select a scenario to see how the dataset shifts.

94
Companies
38%
Win Rate
-10%
Median Return
+89%
Mean Return
Median (bars) and mean (line) across all 6 scenarios
InsightRemoving PLTR barely moves the median (-14% → -16%) but cuts the mean from +96% to +30%. Removing the top 5 returns drops median to -20% — but it was never great. The mean is a lie. The median is the truth. The power law makes these two numbers diverge by 110 percentage points.
Benchmark Comparison

IPO vs Index Benchmarks

Comparing each IPO's return to QQQ and SPY measured from the same IPO date to May 15, 2026. Current benchmark prices: SPY $740.53 · QQQ $710.75 · NASDAQ 26,346.

23%
of IPOs beat QQQ
Only 16 out of 70 companies outperformed QQQ from their IPO date
29%
of IPOs beat SPY
Only 20 out of 70 companies outperformed SPY from their IPO date
-1.06x
median alpha vs QQQ
Median IPO underperformed QQQ by 1.28x — the index wins most of the time
IPO Return vs QQQ vs SPY by Valuation Bucket

Average return from IPO date to today vs what SPY and QQQ returned over the same holding period.

IPO Return
QQQ Return
SPY Return
Mean Alpha vs QQQ by Valuation Bucket

Alpha = IPO return minus QQQ return over the same period. Positive = IPO outperformed the index.

% of IPOs That Beat QQQ vs SPY

How often individual IPOs beat the benchmark from their IPO date to today. 50% line = breakeven.

Beat QQQ %
Beat SPY %
Top 10 by Alpha vs QQQ
TickerBucketAlpha vs QQQBeat?
CRDO$1-2B+20.6xYES
GPCR<$1B+8.3xYES
ATAT$1-2B+7.9xYES
HOOD$7-10B+6.8xYES
ALAB$5-7B+5.0xYES
CRCL$5-7B+3.7xYES
APGE<$1B+3.5xYES
RDDT$5-7B+3.3xYES
NXT$3-5B+3.1xYES
KGS$1-2B+3.1xYES
Bottom 10 by Alpha vs QQQ
TickerBucketAlpha vs QQQBeat?
BMBL$7-10B-2.1xNO
DIDI$40B+-2.0xNO
TASK$10-20B-2.0xNO
COUR$3-5B-2.0xNO
MQ$7-10B-1.9xNO
LZ$7-10B-1.9xNO
BIRD$3-5B-1.8xNO
CRGX<$1B-1.7xNO
UDMY$3-5B-1.7xNO
SG$5-7B-1.7xNO
InsightOnly 23% of IPOs beat QQQ — you'd have been better off buying the index in most cases. The $1–2B bucket is the exception: 50% beat rate and massive alpha driven by CRDO (+20.6x vs QQQ). The $10–20B and $40B+ buckets have the worst beat rates (17%) and deepest negative alpha, confirming the dead zone is real on a risk-adjusted basis too.
Research Findings

12 Insights the Data Actually Supports

1
The $10–20B Dead Zone Is Structural, Not Cyclical
9 companies IPO'd at $10–20B valuations. Only 2 are positive today. Median return: -41%. This pattern holds across 2020 and 2021 cohorts — it persists across different interest rate environments. It's structural: priced too big to re-rate higher, too small to have real pricing power, with no margin for any execution miss.
2
The PLTR Paradox: One Company Rewrites Everything
Palantir (+1415%) accounts for the majority of the dataset's positive mean return. Remove it: mean drops from +96% to +30%. The median barely moves. This is the power law: one company at $21.8B compounding to $330B makes all other analyses secondary.
3
2022 Forced Discipline — and the Best Cohort Emerged
2022 had just 5 IPOs and 60% win rate with +88% median. When the market closed to mediocrity, only genuinely strong companies got through. TPG (+88%), CRDO (+2162%), ATAT (+948%). The IPO market as a filter works when it's closed, not open.
4
2021 Was a Collective Delusion — and the Data Proves It
27 companies IPO'd in 2021. 22 are negative today. Median: -47%. The cohort priced at peak ZIRP multiples (many 40–100× revenue) and has never recovered. Companies like BIRD (-99%), BRZE (-61%), SG (-85%) were great businesses at wrong prices.
5
Semiconductors: 100% Win Rate, No Exception
CRDO (+2162%), ARM (+196%), ATAT (+948%), ACMR (+230%). Every semiconductor IPO in the dataset is positive. The AI chip supercycle isn't selective — it's lifting every company in the compute stack, regardless of size or specific market.
6
The $1–2B Sweet Spot Has the Best Risk-Adjusted Entry
60% win rate, +109% median return, 50% big win rate. These companies are small enough to have 5–10× upside but large enough to have real businesses. CRDO, KGS, LMND, ACMR — the compounders tend to start here.
7
Marketplace Models Are Structurally Broken at IPO
Marketplace/Commerce: 38% win, -55% median. VRM (-99%), WISH (-99%), STUB (-88%), UDMY (-81%). The take-rate business model requires massive capital and scale to work — most never get there, and public market investors eventually re-price the risk.
8
The Power Law Is Extreme: Top 10 Companies Drive Everything
The rank-ordered return curve drops nearly vertically from rank 1 (CRDO, +2162%) to rank 10, then flattens. The top 10 companies produce more total value than the other 84 combined. This isn't venture fund math — it's public market IPO math.
9
Scale Doesn't Protect Mega-IPOs
$40B+ bucket: 29% win rate, -30% median. RIVN (-75%), DiDi (-99%), Venture Global (-69%). Being large doesn't mean having room to grow. ARM (+196%) is the outlier — the only mega-IPO that massively outperformed — driven by AI chip demand nobody priced in at IPO.
10
GoodRx Is the Perfect Dead Zone Case Study
GoodRx: profitable, dominant market position, real revenue, large TAM. IPO'd at $18.9B in 2020. Now worth $1B. -95%. The dead zone doesn't care about your fundamentals when the entry multiple is wrong. Textbook structural trap.
11
2025 Is Already Printing Losses — Despite Selective Market
2025 cohort: 27% win rate, -20% median. StubHub -88%, Gemini -89%, Via -60%. Even in a supposedly post-ZIRP market, companies are overpaying to go public. CRWV (+174%) is the exception. The market is pricing AI premium into everything.
12
The AI Cycle Is Sector-Wide, Not Company-Specific
AI/Data Infra (PLTR, CRWV, C3.ai): 67% win rate, +174% median. Semiconductors: 100% win rate. The AI wave isn't selective between companies — it's creating multiple expansion across every company that touches inference, training, or chip supply. Pick the sector, not just the company.
The Thesis

Valuation at IPO is the variable that matters most

Not the business quality. Not the sector. Not the market environment. The entry valuation — specifically whether you IPO in the $10–20B dead zone or the $1–5B sweet spot — is the single strongest predictor of long-term return in this dataset.

The best compounders (CRDO, PLTR, HOOD, RDDT, ALAB) all entered below $10B. Every single $10–20B IPO has underperformed. The data is unambiguous.

Data: public filings + yfinance (May 2026) · Excludes SPACs, blank-check companies, warrants-only listings · Not investment advice

𝕏 @Trace_Cohent@nyvp.com