Key terms used throughout this page
Reference this if a chart label doesn't make sense on its own.
- Alpha vs QQQ
- A company's return minus what QQQ (the Nasdaq-100 ETF) returned over the same period. +100pp alpha means the stock beat simply buying the index by 100 percentage points; negative means it lagged the index.
- Durability Score
- A 0-100 composite score (the primary ranking metric on this site) that rewards sustained value creation and explicitly penalizes gains concentrated in the 2020-2021 COVID/ZIRP window. 100 = maximally durable; 0 = the floor.
- Archetype
- One of 13 rule-based lifecycle categories describing the shape of a company's outcome -- e.g. "Durable compounder," "Temporary COVID/ZIRP winner," "Permanent capital destroyer." Assigned from the actual price data, not a subjective label.
- Gain Retention
- % of a company's best-ever gain (from IPO price to its peak price) that it still holds today, bounded 0-100%. 100% means the stock is at or above its all-time peak; 0% means it has given back the entire gain.
- Tailwind Penalty
- How much of a company's best market-relative performance came from the Feb 2020-Nov 2021 COVID/ZIRP window specifically, and how much of that advantage has since reversed. High penalty = looked great in 2021, mostly gone now.
- Funding Bucket
- Total venture capital raised before IPO, grouped into 6 bands from under $100M to over $3B. Used throughout the Funding & Valuation section to test whether raising more actually helped.
- Data Quality Tier
- A small badge next to a ticker when its data isn't a plain full daily price history: "Corrected" (a ticker change, acquisition, or bankruptcy was resolved and the price series stitched or terminated), "Synthetic series" (Yahoo purged full daily history for this delisted ticker -- only the verified IPO and terminal prices exist, so only the point-to-point return is reliable), or "Unresolved" (post-listing status couldn't be fully verified). No badge means a full reported daily history.
Did raising more, or IPO'ing at a higher valuation, actually help?
Short answer: no. Companies that raised the least private capital before IPO ($100-250M) show the strongest alpha, while the $3bn+ bucket is the worst. Higher IPO valuations trend toward worse long-term alpha, not better.
Alpha vs QQQ over time, by funding raised
Each line aggregates every company in that funding band at the same point in its life (3 months post-IPO, 6 months, etc.), then plots how that aggregate alpha vs QQQ evolved. A line trending up means that funding cohort kept compounding faster than the market; a line trending toward or below 0pp means it gave back its early advantage. "n=" is the number of companies in that line -- larger n means a steadier, less outlier-driven read. Toggle Mean vs Median above the chart: Mean can be dragged around by a single extreme name (PANW alone pulls its bucket's mean up substantially); Median shows the more typical company instead.
Mean can be skewed by a single extreme outlier (e.g. PANW at +3,650pp). Switch to Median to see the more typical company in each bucket.
Durability & win rate by funding raised
Two views of the same question, side by side per bucket: the solid bar is the median Durability Score (0-100); the faded bar is the win rate -- the % of companies in that bucket that currently beat QQQ at all. If a bucket's two bars are both short, most of its companies underperformed the market outright, not just relative to other IPOs.
IPO valuation vs. current alpha
Did IPO'ing later at a much higher valuation pay off? The trend runs negative — higher valuation correlates with worse current alpha.
Sector × funding raised: median current alpha
Read this like a spreadsheet: pick a sector row and scan across funding-bucket columns to see whether raising more helped or hurt companies specifically in that sector (not the whole market). Sectors with fewer than 4 companies overall are omitted since a single outlier would otherwise dominate the row. Blue cells underperformed QQQ, green cells beat it; deeper color means a larger gap either way. Blank cells simply have no companies in that sector-and-bucket combination.
| Sector | $0–100m | $100–250m | $250–500m | $500m–1bn | $1–3bn | $3bn+ |
|---|---|---|---|---|---|---|
| Consumer Marketplaces | -140pp (n=1) | -339pp (n=1) | -220pp (n=2) | — | -43pp (n=2) | -31pp (n=2) |
| Healthcare Technology | -129pp (n=3) | +361pp (n=2) | -208pp (n=3) | -8pp (n=2) | +127pp (n=2) | — |
| Fintech | -286pp (n=1) | -91pp (n=3) | -140pp (n=5) | -194pp (n=4) | +81pp (n=3) | -128pp (n=1) |
| Data, Analytics, and Databases | — | -246pp (n=3) | -67pp (n=6) | — | +1570pp (n=2) | — |
| Advertising Technology and Media | +375pp (n=2) | -146pp (n=5) | — | -139pp (n=1) | +386pp (n=4) | — |
| Vertical SaaS | -3pp (n=3) | -117pp (n=2) | -102pp (n=2) | -122pp (n=3) | -37pp (n=1) | — |
| Horizontal SaaS | -181pp (n=2) | +33pp (n=2) | -116pp (n=6) | -174pp (n=2) | — | — |
| Other | — | -146pp (n=2) | -317pp (n=1) | -109pp (n=5) | +1100pp (n=1) | — |
| Cloud Infrastructure and Developer Tools | -103pp (n=2) | -63pp (n=5) | -86pp (n=4) | -341pp (n=2) | -33pp (n=2) | — |
| Cybersecurity | +3650pp (n=1) | +284pp (n=3) | +1431pp (n=3) | +68pp (n=2) | — | — |
| Education Technology | — | +101pp (n=2) | -175pp (n=2) | — | — | — |
| E-commerce and Consumer Internet | -268pp (n=1) | -74pp (n=2) | -94pp (n=2) | -98pp (n=2) | -123pp (n=2) | -180pp (n=2) |
What didn't predict durability
"r" is the Pearson correlation coefficient: it runs from -1 (perfectly inverse) to +1 (perfectly aligned), with 0 meaning no relationship at all. Every panel below sits close to 0 (all under 0.15) -- meaning growth rate, margin, Rule of 40, and capital raised at IPO time told you almost nothing about how durable the stock turned out to be. Cohort timing and sector (shown elsewhere on this page) explain far more than the business metrics at IPO did.
Revenue growth at IPO
r = 0.07Gross margin at IPO
r = 0.11Rule of 40 at IPO
r = 0.13Total VC raised ($M)
r = 0.09Initial alpha vs. current alpha
Each bubble is one company. X-axis is alpha vs QQQ in the first 24 months after IPO; Y-axis is alpha vs QQQ as of today. Bubble size is max peak-to-trough drawdown. Upper-right is durable winners; lower-right is early winners that reversed; upper-left is slow starters that became durable winners.
Axes are clipped to the 2nd–98th percentile for readability — 11 extreme outliers (e.g. PANW at +3,650pp current alpha) fall outside this view. Bubble size = max peak-to-trough drawdown. Hover any point, or see the full Company Rankings table below, for exact values and precise archetype.
Returns tell one story. Dollars tell another.
Percentage returns favor smaller companies -- it's easier for a $500M company to 10x than a $100B one. This section translates every company's return into estimated equity value: dollars created from IPO to peak, dollars destroyed since, and net dollars retained today.
Coverage: 93 of 137 companies (68%) have a disclosed IPO valuation and are included above, representing 100% of total known IPO equity value across the dataset. The remaining 44 companies (no disclosed valuation) are excluded from these dollar aggregates rather than estimated.
Percentage return vs. actual equity value created
A high percentage return does not necessarily mean a company created the most total value. PLTR's current return (+1,730%) isn't the highest in this dataset, but at a $15.8B IPO valuation it created roughly $273B in net equity value -- more than the next several companies on the leaderboard below combined.
Largest equity-value creators and destroyers
Ranked in dollars, not percent -- hover any bar for the company's corresponding percentage return, since the two rankings order companies very differently.
Value creation by IPO valuation bucket
The $10B-$25B bucket has a negativemedian return (-38%) yet still generated +$314B in net value, while smaller buckets post stronger median returns on far less total dollar impact. Bigger IPOs don't need a good percentage return to move the most actual money.
| IPO Valuation Bucket | n | Median Return | Total IPO Value | Created to Peak | Destroyed From Peak | Net Value Change |
|---|---|---|---|---|---|---|
| Under $1B | 6 | +158% | $4.1B | $87.5B | -$73.5B | +$14.0B |
| $1B–$5B | 38 | +59% | $86.6B | $653.3B | -$388.4B | +$263.6B |
| $5B–$10B | 22 | +21% | $165.2B | $843.9B | -$426.5B | +$413.4B |
| $10B–$25B | 18 | -38% | $272.8B | $925.7B | -$591.8B | +$314.1B |
| $25B+ | 9 | +59% | $407.2B | $572.2B | -$419.8B | +$152.4B |
Where durability concentrated
Median Conservative Durability Score (0-100) by consolidated sector and by IPO vintage year. The 2021 cohort — the largest single vintage in this dataset at 40 companies — has the weakest median outcome of any year with a meaningful sample size.
By sector (min. 3 companies)
By IPO vintage year
COVID/ZIRP tailwind exposure vs. long-term retention
Companies grouped into 5 bands by tailwind penalty (how much of their best market-relative performance came from the Feb 2020–Nov 2021 window and has since reversed), showing what % of all-time peak gain each band still holds today -- mean (blue) and median (gray) side by side. The median sitting below the mean, and hitting 0% in two bins, shows the mean is partly propped up by a handful of stronger performers rather than reflecting a typical company in that bucket. The middle three bins have few companies (n=4-8) and should be read as noisy either way.
The whole universe, at a glance
Every one of the 132 scored companies, grouped by lifecycle archetype. Size = number of companies. The two largest boxes -- temporary tailwind winners and capital destroyers -- together outnumber every durable-outcome archetype combined.
Full 132-company dataset
Sortable and filterable. Durability is the Conservative Durability Score (0-100, penalizes temporary tailwinds). Alpha figures are in percentage points vs QQQ. Retention and tailwind penalty are bounded 0-100%.
Biggest winners vs. biggest losers, head to head
Ranked by current alpha vs QQQ (not the composite Durability Score, which floors at 0 for dozens of names and loses the story). PANW leads at +3,650pp; GRPN trails at -1,231pp -- a nearly 5,000-percentage-point gap between the best and worst outcome in this dataset.
| Archetype | ||||||
|---|---|---|---|---|---|---|
PANW Palo Alto Networks | 86.9 | +3650pp | +1789pp | 95% | 1% | Durable compounder |
LVGOSynthetic series Livongo | 82.6 | +361pp | — | 100% | 0% | Acquired with positive terminal outcome |
CRDO Credo Technology | 81.8 | +2035pp | +828pp | 73% | 0% | Early rocket, lasting winner |
AVLRSynthetic series Avalara | 81.0 | +232pp | +232pp | 100% | 0% | Acquired with positive terminal outcome |
PSTGCorrected Pure Storage | 80.8 | +328pp | — | 80% | 0% | Recent IPO, insufficient history |
PLANSynthetic series Anaplan | 80.6 | +210pp | +210pp | 100% | 0% | Acquired with positive terminal outcome |
CRWD CrowdStrike | 78.6 | +1842pp | +116pp | 90% | 3% | Early rocket, lasting winner |
HNGE Hinge Health | 78.4 | +127pp | — | 90% | 0% | Recent IPO, insufficient history |
ALAB Astera Labs | 75.6 | +727pp | +153pp | 63% | 0% | Early rocket, lasting winner |
NET Cloudflare | 75.5 | +1431pp | +20pp | 96% | 3% | Early rocket, lasting winner |
SMARSynthetic series Smartsheet | 74.7 | +33pp | +33pp | 100% | 0% | Acquired with positive terminal outcome |
PLTR Palantir | 74.6 | +1570pp | +1360pp | 63% | 7% | Strong, mostly durable |
CBLKSynthetic series Carbon Black | 74.3 | +23pp | — | 100% | 0% | Acquired with positive terminal outcome |
ONEMSynthetic series One Medical | 72.9 | -8pp | -8pp | 100% | 0% | Acquired with positive terminal outcome |
BE Bloom Energy | 71.8 | +1100pp | +1051pp | 64% | 2% | Slow start, long-term winner |
RDDT Reddit | 71.2 | +386pp | +11pp | 64% | 0% | Early rocket, lasting winner |
META Facebook | 70.7 | +457pp | +173pp | 81% | 19% | Durable compounder |
DDOG Datadog | 70.3 | +560pp | -17pp | 91% | 9% | Durable compounder |
MDLASynthetic series Medallia | 69.1 | -43pp | -43pp | 100% | 0% | Acquired with positive terminal outcome |
APP AppLovin | 68.5 | +318pp | +2378pp | 53% | 2% | Slow start, long-term winner |
RBRK Rubrik | 68.3 | +68pp | +38pp | 65% | 0% | Strong, mostly durable |
BASESynthetic series Couchbase | 67.2 | -66pp | -66pp | 100% | 0% | Acquired with positive terminal outcome |
CART Instacart (Maplebear) | 64.5 | -43pp | -20pp | 67% | 0% | Cycle victim |
CRWV CoreWeave | 59.9 | +48pp | — | 28% | 0% | Recent IPO, insufficient history |
OMDA Omada Health | 59.6 | -9pp | — | 51% | 0% | Recent IPO, insufficient history |
PVTLSynthetic series Pivotal Software | 58.9 | -33pp | — | 0% | 0% | Acquired with negative terminal outcome |
CAVA Cava Group | 58.4 | +94pp | -50pp | 32% | 0% | Cycle victim |
HOOD Robinhood Markets | 58.1 | +81pp | +647pp | 60% | 24% | Slow start, long-term winner |
APPF AppFolio | 57.3 | +687pp | -32pp | 50% | 39% | Strong, mostly durable |
CRCL Circle Internet | 56.9 | +99pp | — | 20% | 0% | Recent IPO, insufficient history |
PSCorrected Pluralsight | 55.9 | +101pp | — | 20% | 24% | Acquired with negative terminal outcome |
IOT Samsara | 53.8 | -27pp | -72pp | 34% | 0% | Consistent moderate compounder |
LASR nLight | 52.3 | +9pp | +182pp | 86% | 46% | Slow start, long-term winner |
TTAN ServiceTitan | 49.9 | -37pp | — | 3% | 0% | Recent IPO, insufficient history |
WISHCorrected ContextLogic (Wish) | 46.5 | -123pp | +8pp | 0% | 0% | Cycle victim |
CFLTSynthetic series Confluent | 44.2 | -91pp | -91pp | 0% | 0% | Acquired with negative terminal outcome |
SUMOSynthetic series Sumo Logic | 43.9 | -67pp | -67pp | 0% | 0% | Acquired with negative terminal outcome |
VRMCorrected Vroom | 43.5 | -98pp | — | 0% | 0% | Recent IPO, insufficient history |
DOCN DigitalOcean | 43.1 | +56pp | +156pp | 67% | 63% | Slow start, long-term winner |
MDB MongoDB | 41.7 | +789pp | -131pp | 51% | 64% | Early rocket, full reversal |
FIG Figma | 41.5 | -62pp | — | 0% | 0% | Recent IPO, insufficient history |
CHME Chime | 41.4 | -128pp | — | 0% | 0% | Recent IPO, insufficient history |
UBER Uber | 41.4 | -242pp | -67pp | 48% | 0% | Cycle victim |
POSHSynthetic series Poshmark | 41.0 | -74pp | -74pp | 0% | 0% | Acquired with negative terminal outcome |
ZS Zscaler | 40.8 | +492pp | -28pp | 38% | 76% | Cycle victim |
SVMKSynthetic series SurveyMonkey | 40.4 | -116pp | -116pp | 0% | 0% | Acquired with negative terminal outcome |
CSPRSynthetic series Casper Sleep | 40.2 | -94pp | — | 0% | 0% | Acquired with negative terminal outcome |
OSCR Oscar Health | 39.3 | -158pp | +379pp | 0% | 0% | Slow start, long-term winner |
HCPSynthetic series HashiCorp | 39.3 | -86pp | -86pp | 0% | 0% | Acquired with negative terminal outcome |
KVYO Klaviyo | 39.0 | -139pp | -69pp | 0% | 0% | Cycle victim |
MNTN MNTN | 38.3 | -81pp | — | 0% | 0% | Recent IPO, insufficient history |
TWLO Twilio | 36.4 | +605pp | -84pp | 42% | 77% | Early rocket, full reversal |
DUOL Duolingo | 35.5 | -77pp | -105pp | 5% | 25% | Cycle victim |
BIGCCorrected BigCommerce | 35.4 | -117pp | — | 0% | 0% | Recent IPO, insufficient history |
IBTA Ibotta | 35.3 | -140pp | -27pp | 0% | 0% | Cycle victim |
ZICorrected ZoomInfo | 35.1 | -102pp | — | 0% | 0% | Recent IPO, insufficient history |
VEEV Veeva Systems | 35.0 | -3pp | +70pp | 53% | 100% | Temporary COVID/ZIRP winner |
ROOT Root | 34.2 | -258pp | +501pp | 0% | 0% | Permanent capital destroyer |
YOU Clear Secure | 32.2 | -23pp | +78pp | 83% | 100% | Temporary COVID/ZIRP winner |
ACCDSynthetic series Accolade | 30.7 | -138pp | -138pp | 0% | 0% | Acquired with negative terminal outcome |
FROG JFrog | 29.3 | -63pp | +120pp | 79% | 100% | Temporary COVID/ZIRP winner |
OLOSynthetic series Olo | 28.9 | -146pp | -146pp | 0% | 0% | Acquired with negative terminal outcome |
DASH DoorDash | 27.5 | -59pp | +83pp | 48% | 100% | Temporary COVID/ZIRP winner |
BOX Box | 23.8 | -543pp | -428pp | 65% | 0% | Cycle victim |
OKTA Okta | 23.3 | +284pp | -254pp | 45% | 81% | Early rocket, full reversal |
ABNB Airbnb | 23.1 | -31pp | -105pp | 51% | 100% | Temporary COVID/ZIRP winner |
UDMYSynthetic series Udemy | 23.1 | -175pp | -175pp | 0% | 0% | Acquired with negative terminal outcome |
COMP Compass | 22.8 | -162pp | +128pp | 0% | 48% | Slow start, long-term winner |
NTNX Nutanix | 22.6 | -301pp | -275pp | 57% | 13% | Early rocket, full reversal |
DBX Dropbox | 22.2 | -341pp | -239pp | 40% | 0% | Cycle victim |
TTD Trade Desk | 21.6 | +375pp | -269pp | 12% | 76% | Early rocket, full reversal |
SNOW Snowflake | 21.6 | -41pp | -104pp | 54% | 100% | Temporary COVID/ZIRP winner |
LMND Lemonade | 21.6 | -54pp | +109pp | 26% | 100% | Temporary COVID/ZIRP winner |
AFRM Affirm | 21.3 | -79pp | +323pp | 22% | 100% | Temporary COVID/ZIRP winner |
ZUOSynthetic series Zuora | 19.2 | -276pp | -276pp | 0% | 0% | Acquired with negative terminal outcome |
HUBS HubSpot | 17.5 | +113pp | -211pp | 23% | 96% | Cycle victim |
SDCSynthetic series SmileDirectClub | 17.5 | -208pp | -208pp | 0% | 0% | Delisted or bankrupt |
RENT Rent the Runway | 17.3 | -191pp | -180pp | 0% | 0% | Permanent capital destroyer |
REAL The RealReal | 15.9 | -336pp | -153pp | 0% | 0% | Cycle victim |
RELY Remitly Global | 15.3 | -140pp | -105pp | 0% | 51% | Cycle victim |
YELP Yelp | 11.8 | -1065pp | -836pp | 13% | 0% | Early rocket, full reversal |
COIN Coinbase | 11.5 | -147pp | +135pp | 0% | 100% | Temporary COVID/ZIRP winner |
UPST Upstart | 11.2 | -91pp | -74pp | 3% | 100% | Temporary COVID/ZIRP winner |
EBSynthetic series Eventbrite | 10.6 | -326pp | -326pp | 0% | 0% | Acquired with negative terminal outcome |
PGNY Progyny | 10.3 | -129pp | -144pp | 36% | 100% | Temporary COVID/ZIRP winner |
LYFT Lyft | 10.2 | -391pp | -208pp | 0% | 0% | Permanent capital destroyer |
BLND Blend Labs | 10.0 | -194pp | -34pp | 0% | 64% | Permanent capital destroyer |
BLZE Backblaze | 9.8 | -103pp | +44pp | 0% | 100% | Temporary COVID/ZIRP winner |
YEXT Yext | 8.2 | -502pp | -368pp | 0% | 0% | Cycle victim |
TOST Toast | 8.1 | -122pp | -38pp | 0% | 100% | Temporary COVID/ZIRP winner |
RBLX Roblox | 7.4 | -109pp | -108pp | 8% | 100% | Temporary COVID/ZIRP winner |
DOCS Doximity | 4.1 | -128pp | -133pp | 0% | 100% | Temporary COVID/ZIRP winner |
TENB Tenable | 3.0 | -266pp | -173pp | 28% | 88% | Temporary COVID/ZIRP winner |
PD PagerDuty | 2.9 | -361pp | -195pp | 0% | 37% | Cycle victim |
GRPN Groupon | 2.9 | -1231pp | -903pp | 0% | 0% | Permanent capital destroyer |
MQ Marqeta | 2.9 | -201pp | -112pp | 0% | 67% | Permanent capital destroyer |
S SentinelOne | 2.0 | -152pp | -71pp | 0% | 100% | Temporary COVID/ZIRP winner |
BILL Bill.com | 1.3 | -156pp | -166pp | 7% | 100% | Temporary COVID/ZIRP winner |
PCOR Procore | 0.6 | -156pp | -138pp | 0% | 100% | Temporary COVID/ZIRP winner |
ZM Zoom | 0.4 | -146pp | -187pp | 10% | 100% | Temporary COVID/ZIRP winner |
NRDS NerdWallet | 0.2 | -133pp | -116pp | 0% | 100% | Temporary COVID/ZIRP winner |
AI C3.ai | 0.0 | -223pp | -187pp | 0% | 100% | Permanent capital destroyer |
AMPL Amplitude | 0.0 | -175pp | -118pp | 0% | 100% | Permanent capital destroyer |
AMWL Amwell | 0.0 | -268pp | -235pp | 0% | 100% | Permanent capital destroyer |
APPN Appian | 0.0 | -320pp | -356pp | 6% | 100% | Temporary COVID/ZIRP winner |
ASAN Asana | 0.0 | -225pp | -238pp | 0% | 100% | Temporary COVID/ZIRP winner |
BMBL Bumble | 0.0 | -220pp | -237pp | 0% | 100% | Permanent capital destroyer |
BRZE Braze | 0.0 | -146pp | -138pp | 0% | 100% | Temporary COVID/ZIRP winner |
BYND Beyond Meat | 0.0 | -392pp | -216pp | 0% | 75% | Permanent capital destroyer |
CDLX Cardlytics | 0.0 | -477pp | -321pp | 0% | 100% | Permanent capital destroyer |
COUR Coursera | 0.0 | -213pp | -179pp | 0% | 100% | Permanent capital destroyer |
CPNG Coupang | 0.0 | -180pp | -127pp | 0% | 100% | Temporary COVID/ZIRP winner |
CXM Sprinklr | 0.0 | -174pp | -156pp | 0% | 100% | Temporary COVID/ZIRP winner |
DIBS 1stDibs | 0.0 | -192pp | -89pp | 0% | 100% | Permanent capital destroyer |
DOCU DocuSign | 0.0 | -286pp | -295pp | 8% | 100% | Temporary COVID/ZIRP winner |
ESTC Elastic | 0.0 | -246pp | -213pp | 16% | 100% | Temporary COVID/ZIRP winner |
EVER EverQuote | 0.0 | -286pp | -258pp | 20% | 100% | Temporary COVID/ZIRP winner |
EXFY Expensify | 0.0 | -181pp | -106pp | 0% | 95% | Permanent capital destroyer |
FLYW Flywire | 0.0 | -146pp | -148pp | 0% | 100% | Temporary COVID/ZIRP winner |
FRSH Freshworks | 0.0 | -168pp | -146pp | 0% | 100% | Permanent capital destroyer |
FSLY Fastly | 0.0 | -275pp | -175pp | 4% | 100% | Temporary COVID/ZIRP winner |
GTLB GitLab | 0.0 | -156pp | -125pp | 0% | 100% | Temporary COVID/ZIRP winner |
HCAT Health Catalyst | 0.0 | -375pp | -195pp | 0% | 54% | Permanent capital destroyer |
LZ LegalZoom | 0.0 | -180pp | -135pp | 0% | 100% | Permanent capital destroyer |
NCNO nCino | 0.0 | -226pp | -195pp | 0% | 100% | Temporary COVID/ZIRP winner |
PATH UiPath | 0.0 | -194pp | -150pp | 0% | 100% | Permanent capital destroyer |
PHR Phreesia | 0.0 | -325pp | -188pp | 0% | 100% | Temporary COVID/ZIRP winner |
PINS Pinterest | 0.0 | -275pp | -184pp | 5% | 100% | Temporary COVID/ZIRP winner |
PTON Peloton | 0.0 | -368pp | -189pp | 0% | 100% | Permanent capital destroyer |
RIVN Rivian Automotive | 0.0 | -165pp | -75pp | 0% | 100% | Permanent capital destroyer |
RNG RingCentral | 0.0 | -682pp | -527pp | 6% | 100% | Temporary COVID/ZIRP winner |
RVLV Revolve | 0.0 | -268pp | -173pp | 11% | 100% | Temporary COVID/ZIRP winner |
SONO Sonos | 0.0 | -317pp | -184pp | 0% | 100% | Temporary COVID/ZIRP winner |
SPT Sprout Social | 0.0 | -308pp | -175pp | 0% | 100% | Temporary COVID/ZIRP winner |
U Unity | 0.0 | -217pp | -167pp | 0% | 100% | Temporary COVID/ZIRP winner |
UPWK Upwork | 0.0 | -339pp | -219pp | 0% | 100% | Temporary COVID/ZIRP winner |
WDAY Workday | 0.0 | -706pp | -616pp | 41% | 92% | Temporary COVID/ZIRP winner |
Z Zillow | 0.0 | -601pp | -434pp | 7% | 100% | Temporary COVID/ZIRP winner |
Company lifecycles, in daily detail
Weekly-resolution cumulative return and alpha vs QQQ since IPO for seven validation names spanning every archetype in this research.
Huge 24-month alpha, then a near-total reversal — a textbook temporary ZIRP winner.
Seven conclusions this data supports
These are inferences drawn directly from the 132-company dataset above, not assumptions that went in before the analysis — several of them (fundamentals not predicting durability, higher valuations trending worse) run against what the original research thesis expected to find. Most are based on simple correlations and raw-sample comparisons, not controlled regressions -- treat the associations below as leads worth investigating further, not settled causal claims.
Public-market value from IPOs is extremely concentrated
Only 24% of companies (32 of 132) show any positive alpha vs. QQQ since their IPO offer price at all. Of those, the top 10 winners alone account for 76% of all the positive alpha generated across the entire universe. Most of what looks like an “IPO asset class” is really a small number of extreme winners subsidizing a much larger set of laggards and destroyers.
IPO vintage showed a stronger relationship with outcomes than the fundamentals we measured
The 2021 IPO vintage — the largest single cohort in this dataset at 40 companies — has a 7.5% win rate and a median Durability Score of 7.5 out of 100, the worst outcome of any year with a meaningful sample size. Meanwhile revenue growth, gross margin, Rule of 40, and capital raised at IPO all show near-zero simple correlation (r under 0.15) with long-term durability. That's a univariate read, not a controlled one -- it doesn't rule out fundamentals mattering within a given sector or vintage, only that they didn't show a signal across the raw sample.
Higher pre-IPO capital raised was associated with weaker outcomes in this sample
Companies that raised the least private capital before IPO ($100-250M) show the strongest average alpha; the $3B+ bucket is the worst performer of any funding band. We haven't controlled for vintage or sector here, and companies that raise $3B+ are likely structurally different from $100M raisers in ways beyond the funding amount itself -- so treat this as an association worth investigating, not an isolated causal effect of capital raised.
Higher IPO valuations trended toward worse outcomes, not better
Plotting IPO valuation against current alpha shows a negative relationship: companies that stayed private longer to reach a much higher valuation (COIN, CPNG, HOOD, ABNB — all in the $50B+ range) cluster near flat-to-negative alpha, while some of the cheapest IPOs in the dataset (APPF at $310M, PANW, MDB) delivered the biggest long-term wins. One extreme outlier (PANW) steepens this trend considerably -- the direction holds without it, but more weakly (see the chart's own caption for the exact slope with and without it).
A third of the universe rode a tailwind that has since reversed
43 of 132 companies (33%) are classified as “Temporary COVID/ZIRP winners” — their best market-relative performance was concentrated in the Feb 2020–Nov 2021 window and has substantially given it back since. This is a far longer list than the household names (BILL, ZM) usually cited as COVID winners.
Early performance is a weak predictor — slow starters can still win big
7 companies in this dataset (APP, PLTR, BE, and others) had weak or negative returns in their first 24 months, then went on to deliver some of the strongest long-term compounding in the universe. Writing off a company based on its first two years post-IPO would have missed these entirely.
Data-quality issues were common enough to change specific conclusions
This research caught two issues that would have materially misranked individual companies if left unaddressed: a split-adjustment mismatch that turned PANW's true +3,650pp alpha into a false -422pp reading, and Yahoo Finance silently purging full price history for many acquired or delisted tickers. Any comparable analysis that doesn't explicitly check for these will misprice some of its highest-conviction names.
Built on daily prices, not checkpoints — with the data-quality issues we found corrected along the way
Split-adjustment mismatch
yfinance's adjusted close is post-split; the source workbook's IPO offer price is not. Before fixing this, PANW (two splits, 6x cumulative) showed a false -422pp alpha. After adjustment: +3,650pp — now the #1 ranked durable compounder.
Delisted-ticker data gaps
Yahoo Finance purges full price history for many acquired/delisted tickers, not just post-delisting dates. 20 companies use a verified synthetic IPO-price → terminal-value series instead of daily history (flagged with a "Synthetic series" badge in the rankings table); every ticker in the universe now has a verified resolution.
Bounded, price-space retention
Gain retention is computed as (current − IPO price) / (peak − IPO price), clipped to 0–100%, in price space rather than an unbounded alpha ratio — avoiding nonsensical outputs like -46x from a near-zero early-gain denominator.
What surprised the analysis
A controlled regression — demeaning both outcomes and fundamentals within each IPO-year x sector cohort, so it isn't just picking up "2021 SaaS did badly" — still finds no significant relationship between at-IPO fundamentals (revenue growth, gross margin, Rule of 40, revenue scale, profitability) and long-term alpha (n=67, R²=0.19, none of the 5 fundamentals significant at 5%). Cohort timing and sector still dominate.
Regime calendar, cross-checked
The 7 market-regime windows (COVID crash, ZIRP boom, etc.) are hand-picked from known events, not derived from price data — so we checked each boundary date against QQQ's own peaks/troughs. The two sharpest, most consequential turns (the Feb 2020 COVID crash and the Nov 2021 ZIRP-boom peak) land within a day of a real QQQ extremum; the softer, calendar-driven boundaries (year-end cutoffs) sit 3-4 weeks off a true turning point, as expected for a boundary chosen for readability rather than a price signal.