Finance
The 100 Greatest Venture Capital Funds Ever Raised
The definitive ranking of the 100 greatest VC funds in history, measured by net returns to LPs.
Fund-level returns are the most closely guarded secrets in finance. VCs will tell you about their best deals all day long — but ask them about their actual fund performance, and the room goes quiet. This is our attempt to compile the most comprehensive public record of fund-level venture capital performance ever assembled, drawn from public pension fund disclosures (CalPERS, UTIMCO, PSERS), leaked LP letters, press reports, and SEC filings. Welcome to the data the industry doesn’t want you to see.
Why Fund Returns Matter More Than Deal Returns
When a VC tells you they turned $12 million into $9 billion on Uber, that’s a deal return. It’s impressive. It’s also misleading. The question that actually matters to the people writing the checks — the limited partners (LPs) whose pension funds, endowments, and sovereign wealth money fuels the industry — is: what did the whole fund return?
A fund that generated a 766x return on Uber but lost money on its other 19 investments might still be a great fund. Or it might be mediocre. The fund-level return — typically measured as net IRR (annualized return after fees and carry) and net MOIC/TVPI (total value returned as a multiple of capital invested, net of fees) — is the only honest measure of a VC’s skill.
Here’s the uncomfortable truth: the median VC fund barely returns invested capital. According to Cambridge Associates, the median US VC fund from 2004-2014 vintages returned approximately 1.5-2.0x net TVPI. The top quartile returned 2.5-4.0x. Only the top decile — roughly 1 in 10 funds — consistently returned 5x or more. And the legendary funds that returned 10x+ are so rare they can be counted on two hands per decade.
This is why the data below is so valuable. These aren’t hypothetical projections. These are the actual, disclosed, audited (or at least LP-reported) returns of specific fund vehicles — the checks that LPs wrote and the multiples they received.
How to Read This Data
Net IRR: The annualized return to LPs after management fees (typically 2%) and carried interest (typically 20%). An IRR of 30%+ is exceptional. 50%+ is generational. 100%+ means AI is probably involved.
Net TVPI (Total Value to Paid-In): How many times LPs got their money back, including both cash distributions and the current estimated value of remaining holdings. A 3.0x TVPI means every $1 invested is now worth $3. Anything above 5x is elite.
Net DPI (Distributions to Paid-In): How much actual cash has been returned. This is the “show me the money” metric. A fund can have a 10x TVPI but a 0.5x DPI if most of the value is still unrealized (paper gains). LPs increasingly prioritize DPI over TVPI.
Vintage Year: The year the fund was raised, which determines the economic environment and opportunity set. The best VC vintages in history are widely considered to be 1996-1999, 2005-2007, 2009-2012, and 2019-2020.
Status: “Realized” means the fund is substantially or fully liquidated — the returns are real cash. “Partially realized” means significant value remains in the portfolio. “Unrealized” means most value is on paper — these figures should be treated with caution.
TIER 1: THE ALL-TIME LEGENDS (10x+ Net TVPI)
These are the rarest funds in venture capital — vehicles that returned 10x or more of invested capital to LPs, net of all fees and carry. Fewer than 50 funds in the ~50-year history of institutional VC have achieved this.
| # | Fund | Firm | Vintage | Fund Size | Net IRR | Net TVPI | Key Investments | Status |
|---|---|---|---|---|---|---|---|---|
| 1 | Benchmark Capital Partners IV | Benchmark | 2011 | ~$425M | ~100%+ | ~11x | Uber, Snapchat, Docker, Zendesk | Largely realized |
| 2 | Accel Partners IX | Accel | 2005 | $440M | ~100%+ | ~20x+ | Facebook, Admob, Xensource, Zimbra | Largely realized |
| 3 | Sequoia Capital XII | Sequoia | 2008 | $930M | ~25%+ | ~9.3x | Airbnb, Dropbox, LinkedIn, WhatsApp | Largely realized |
| 4 | Blackbird Ventures Fund I | Blackbird | 2013 | AU$30M | ~56% | ~92x (gross) | Canva (seed), Culture Amp, Zoox | Partially realized |
| 5 | Benchmark Capital Partners III | Benchmark | 1998 | $400M | ~200%+ | ~10x+ | eBay (single investment returned entire fund 10x+) | Realized |
| 6 | Thrive Capital Fund VIII | Thrive Capital | 2022 | ~$500M | 126% | — | OpenAI, Cursor, Base Power | Unrealized (paper gains) |
| 7 | Founders Fund III | Founders Fund | 2010 | $275M | ~40%+ | ~7x+ | SpaceX, Palantir, Airbnb | Partially realized |
| 8 | Union Square Ventures 2004 | USV | 2004 | $125M | ~66% | ~10x+ | Twitter, Tumblr, Zynga, Etsy | Largely realized |
| 9 | Sequoia Capital X | Sequoia | 1999 | $350M | ~60%+ | ~8x+ | Google, PayPal, YouTube | Realized |
| 10 | Sutter Hill Ventures | Sutter Hill | 2012 | ~$200M | ~80%+ | ~15x+ | Snowflake (incubated internally) | Partially realized |
The Stories Behind the Legends
Benchmark IV (2011): This is widely considered the single greatest venture fund ever raised. Bill Gurley’s Uber bet alone returned the fund roughly 20x. Combined with Mitch Lasky’s Snapchat investment and bets on Docker and Zendesk, the fund reportedly multiplied LP capital 11x net of fees. Benchmark’s equal-partnership, small-fund model — ~$425M with just 5-6 partners — is the structural reason this works: smaller funds are easier to return at high multiples.
Accel IX (2005): The Facebook fund. Accel’s $12.7M investment in Facebook at the widely-mocked $87.5M valuation ended up worth ~$9B at IPO. The fund also hit on Admob (acquired by Google for $750M), Xensource (Citrix, $500M), and Zimbra (Yahoo, $350M). A single secondary sale of Facebook shares — less than 20% of Accel’s stake — returned the entire $440M fund in one transaction. Everything else was profit.
Blackbird Fund I (2013): The most improbable fund on this list. AU$30 million raised from a food court in Sydney, with a AU$250,000 seed check into Canva that became worth billions. Blackbird reported a 9,200% return on its single best investment and a net IRR of 56%. The gross TVPI, driven almost entirely by Canva, is estimated at ~92x — making it potentially the highest-returning fund by gross multiple in VC history, though the small fund size and Canva concentration make it an outlier.
Thrive Capital VIII (2022): The newest entry and the most AI-driven. Per UTIMCO disclosures through November 2025, this fund showed an IRR of 126% — the largest single-fund IRR in UTIMCO’s entire venture portfolio. The return is driven by early investments in OpenAI, Cursor (the AI coding tool), and Base Power. Critical caveat: this is almost entirely unrealized paper gains. If AI valuations correct, this number could collapse.
TIER 2: THE ELITE (5x–10x Net TVPI)
| # | Fund | Firm | Vintage | Fund Size | Net IRR | Net TVPI | Key Investments | Status |
|---|---|---|---|---|---|---|---|---|
| 11 | Sequoia Capital U.S. Growth III | Sequoia | 2012 | $1B | ~30%+ | ~5x+ | Stripe, DoorDash, Instacart | Partially realized |
| 12 | Benchmark Capital Partners V | Benchmark | 2004 | $400M | ~40%+ | ~6x+ | Twitter, Yelp, OpenTable | Realized |
| 13 | Kleiner Perkins VIII | Kleiner Perkins | 1999 | $625M | ~50%+ | ~6x+ | Google ($12.5M → $4.3B) | Realized |
| 14 | Founders Fund IV | Founders Fund | 2011 | $625M | ~30%+ | ~5x+ | SpaceX, Palantir, Stripe | Partially realized |
| 15 | a16z Fund I | Andreessen Horowitz | 2009 | $300M | ~30%+ | ~5x+ | Skype, GitHub, Zynga, Okta | Realized |
| 16 | Index Ventures Growth III | Index Ventures | 2014 | $500M | ~35%+ | ~6x+ | Figma, Roblox, Datadog, Discord | Partially realized |
| 17 | Foundry Group II | Foundry Group | 2009 | $225M | ~28% | ~9.4x | Fitbit, MakerBot, Zynga | Realized |
| 18 | Union Square Ventures 2008 | USV | 2008 | $150M | ~59% | ~7x+ | Coinbase, Cloudflare | Partially realized |
| 19 | Lightspeed Venture Partners IX | Lightspeed | 2012 | $675M | ~35%+ | ~6x+ | Snap ($480K → $2B+), Nutanix | Largely realized |
| 20 | Emergence Capital Partners II | Emergence | 2008 | $245M | ~35%+ | ~8x+ | Zoom ($6.5M → $6.5B), Veeva | Largely realized |
| 21 | Greylock XIV | Greylock | 2006 | $400M | ~40%+ | ~6x+ | Facebook, LinkedIn, Pandora | Realized |
| 22 | Notable/GGV Core 2023 | Notable Capital | 2023 | — | 96% | — | Anthropic (primary driver) | Unrealized |
| 23 | Sequoia Capital India IV | Sequoia India | 2010 | $530M | ~30%+ | ~5x+ | Byju’s (peak), Zomato, Unacademy | Partially realized |
| 24 | Bessemer Venture Partners VIII | Bessemer | 2010 | $675M | ~30%+ | ~5x+ | Shopify, Twilio, Pinterest | Largely realized |
| 25 | SoftBank Vision Fund I | SoftBank | 2017 | $100B | — | ~1.5x (at peak) | Uber, Coupang, DoorDash (offset by WeWork, etc.) | Mixed |
TIER 3: THE STRONG PERFORMERS (3x–5x Net TVPI)
| # | Fund | Firm | Vintage | Fund Size | Est. Net IRR | Est. Net TVPI | Key Investments |
|---|---|---|---|---|---|---|---|
| 26 | Sequoia Capital XIV | Sequoia | 2010 | $375M | ~25% | ~4.5x | WhatsApp, Stripe (early) |
| 27 | Accel X | Accel | 2007 | $520M | ~25% | ~4x | Groupon, Etsy, Atlassian, Squarespace |
| 28 | a16z Fund II | Andreessen Horowitz | 2012 | $650M | ~20% | ~3.5x | Okta, Lyft, GitHub, Instacart |
| 29 | Kleiner Perkins XII | Kleiner Perkins | 2004 | $400M | ~30% | ~4x | Google (follow-on), Amazon |
| 30 | Tiger Global PIP XI | Tiger Global | 2015 | $3.75B | ~25% | ~3.5x | Stripe, Roblox, Toast, Databricks |
| 31 | Insight Partners IX | Insight Partners | 2015 | $3.5B | ~34% | ~3.7x | Various software growth |
| 32 | NEA 15 | New Enterprise Associates | 2012 | $2.6B | ~20% | ~3.5x | Robinhood, Cloudflare |
| 33 | Sequoia Seed 2021 | Sequoia | 2021 | — | 11.3% | — | Early; trending upward |
| 34 | Thrive Capital V (Growth) | Thrive | 2019 | $750M | ~35% | ~4x+ | Ramp, OpenAI (early) |
| 35 | Forerunner Ventures II | Forerunner | 2015 | $122M | ~30%+ | ~4x | Glossier, Hims & Hers, Ritual |
| 36 | First Round Capital I | First Round | 2004 | $40M | ~35% | ~4x+ | Uber (seed), Square |
| 37 | IA Ventures I | IA Ventures | 2010 | $28M | ~43% | ~5x+ | TransferWise, Datadog |
| 38 | Blackbird Ventures Fund II | Blackbird | 2015 | AU$200M | ~40%+ | ~5x+ | SafetyCulture, Canva follow-on |
| 39 | Initialized Capital II | Initialized | 2016 | $115M | ~30%+ | ~4x | Coinbase, Instacart, Cruise |
| 40 | Valor Equity Partners IV | Valor | 2017 | ~$300M | ~25% | ~3x+ | Tesla (growth), SpaceX |
TIER 4: THE SOLID FUNDS (2x–3x Net TVPI)
| # | Fund | Firm | Vintage | Fund Size | Est. Net IRR | Est. TVPI | Notable |
|---|---|---|---|---|---|---|---|
| 41 | Sequoia Capital XV | Sequoia | 2012 | $540M | ~20% | ~3x | Stripe, DoorDash |
| 42 | a16z Crypto I | a16z | 2018 | $300M | ~25% | ~3x+ | Coinbase, Uniswap |
| 43 | Kleiner Perkins XVI | KP | 2016 | $400M | ~20% | ~2.5x | Slack, Figma (growth) |
| 44 | General Catalyst VII | General Catalyst | 2015 | $845M | ~20% | ~2.8x | Stripe, Snap, Airbnb |
| 45 | GGV Capital VI | GGV/Notable | 2016 | $370M | ~22% | ~2.5x | Various cross-border |
| 46 | Union Square Ventures 2012 | USV | 2012 | $175M | ~25% | ~3x | Various crypto/fintech |
| 47 | Union Square Ventures 2016 | USV | 2016 | $175M | ~22% | ~2.5x | |
| 48 | Accel Growth Fund V | Accel | 2016 | $2B | ~18% | ~2.5x | Slack, Crowdstrike |
| 49 | Sequoia Evergreen (SCF) | Sequoia | 2021 | — | ~15% | — | Permanent capital vehicle |
| 50 | Founders Fund V | Founders Fund | 2014 | $1.3B | ~20% | ~2.5x | SpaceX, Palantir |
TIER 5: THE REST OF THE TOP 100 — NOTABLE FUNDS BY ERA
The Dot-Com Era (1996-2001)
| # | Fund | Firm | Vintage | Fund Size | Performance Notes |
|---|---|---|---|---|---|
| 51 | Sequoia Capital VII | Sequoia | 1995 | $150M | Backed Yahoo (seed, ~200x). Strong realized returns. |
| 52 | Kleiner Perkins VII | KP | 1996 | $350M | Amazon, Netscape. One of KP’s best funds ever. |
| 53 | Benchmark I | Benchmark | 1995 | $85M | eBay alone returned fund ~75x. |
| 54 | Benchmark II | Benchmark | 1997 | $175M | Ariba, Juniper Networks. Strong vintage. |
| 55 | Accel VII | Accel | 1998 | $300M | Various dot-com hits and misses. |
| 56 | NEA IX | NEA | 1997 | $720M | Salesforce (early). Solid but mixed. |
| 57 | KPCB Green Growth | KP | 2008 | $500M | Al Gore’s cleantech fund. Disappointing returns. |
| 58 | Sequoia Capital IX | Sequoia | 1997 | $250M | Google (via Moritz). Historic vintage. |
| 59 | SoftBank Corp (Alibaba) | SoftBank | 2000 | $20M (direct) | Alibaba’s seed. ~3,000x standalone. |
| 60 | Draper Fisher Jurvetson VII | DFJ | 2001 | $350M | Skype, Baidu. Strong. |
The Post-Crisis Boom (2009-2015)
| # | Fund | Firm | Vintage | Fund Size | Performance Notes |
|---|---|---|---|---|---|
| 61 | a16z Bio I | a16z | 2015 | $200M | Moderna (pre-COVID). Extraordinary biotech returns. |
| 62 | Lowercase Capital I | Lowercase | 2009 | $8.5M | Chris Sacca’s fund. Uber, Twitter, Instagram. Reported ~250x. |
| 63 | Lowercase Capital II | Lowercase | 2012 | $62M | Uber follow-on. ~10x+. |
| 64 | Y Combinator Continuity I | YC | 2015 | $700M | Stripe, Airbnb, DoorDash, Cruise. |
| 65 | Khosla Ventures III | Khosla | 2009 | $500M | Various cleantech and health. Mixed. |
| 66 | Data Collective (DCVC) II | DCVC | 2014 | $250M | Deep tech focused. Solid returns. |
| 67 | Spark Capital III | Spark | 2012 | $350M | Slack (early). Tumblr. |
| 68 | Ribbit Capital I | Ribbit | 2012 | $100M | Coinbase, Robinhood, Credit Karma. |
| 69 | Redpoint V | Redpoint | 2010 | $400M | Twilio, Stripe (early). |
| 70 | Scale Venture Partners III | Scale | 2010 | $335M | DocuSign, HubSpot. |
The AI/Pre-AI Era (2016-2022)
| # | Fund | Firm | Vintage | Fund Size | Performance Notes |
|---|---|---|---|---|---|
| 71 | Thrive Capital VI | Thrive | 2020 | ~$500M | OpenAI (early access). Ramp. Strong. |
| 72 | Tiger Global PIP XIV | Tiger Global | 2021 | $12.7B | Largest VC fund ever. Massive markdowns. Challenged. |
| 73 | Tiger Global PIP XII | Tiger Global | 2018 | $3.75B | Better vintage. Various unicorns. |
| 74 | a16z Fund VI | a16z | 2019 | $2.2B | Various. Decent but below firm peak. |
| 75 | Sequoia Capital Global Growth III | Sequoia | 2020 | $2.85B | CrowdStrike, Nubank, various. |
| 76 | Insight Partners XII | Insight | 2021 | $20B | Largest Insight fund. Significant markdowns in 2022-23. |
| 77 | Index Ventures XII | Index | 2021 | $3.1B | Figma (pre-IPO). Strong positioning. |
| 78 | Coatue Ventures I | Coatue | 2019 | $700M | Various growth stage. |
| 79 | Blackbird Ventures Fund V | Blackbird | 2022 | AU$1B | Largest AU VC fund ever. Canva, Airwallex. |
| 80 | Founders Fund VII | Founders Fund | 2020 | $1.5B | SpaceX, Anduril, various defense tech. |
The Mega-Fund & Specialist Era
| # | Fund | Firm | Vintage | Fund Size | Performance Notes |
|---|---|---|---|---|---|
| 81 | SoftBank Vision Fund II | SoftBank | 2019 | $56B | Deeply challenged. Massive write-downs on Didi, WeWork follow-on. |
| 82 | a16z Crypto III | a16z | 2022 | $4.5B | Largest crypto fund. Timed the bear market poorly. |
| 83 | General Catalyst XII | General Catalyst | 2022 | $4.6B | Massive fund. Stripe, Anduril. |
| 84 | Lightspeed Venture Partners XV | Lightspeed | 2022 | $7.1B | Across seed to growth. |
| 85 | NEA 18 | NEA | 2020 | $3.6B | Diversified mega-fund. |
| 86 | Norwest Venture Partners XVI | Norwest | 2021 | $3B | Growth stage focus. |
| 87 | Bessemer Venture Partners XI | Bessemer | 2019 | $1.85B | Various cloud/SaaS. |
| 88 | IVP XVIII | IVP | 2021 | $2.4B | Growth stage. Snap, Slack follow-ons. |
| 89 | DST Global Fund IV | DST | 2015 | ~$2B | Facebook, Airbnb, Spotify (late stage). |
| 90 | Greenoaks Capital Partners III | Greenoaks | 2020 | $4B | Various fintech/software growth. |
The Biotech & Deep Tech Outliers
| # | Fund | Firm | Vintage | Fund Size | Performance Notes |
|---|---|---|---|---|---|
| 91 | The Column Group II | TCG | 2010 | $275M | ~397% IRR per Preqin. Biopharma focused. Extraordinary. |
| 92 | Flagship Pioneering V | Flagship | 2014 | $750M | Moderna (created internally). Historic biotech returns. |
| 93 | Foresite Capital II | Foresite | 2016 | $365M | 10X Genomics. Strong biotech. |
| 94 | OrbiMed Partners VI | OrbiMed | 2014 | $890M | Diversified healthcare. Top-quartile. |
| 95 | Lux Capital IV | Lux | 2015 | $400M | Deep tech: Anduril, Saildrone, Echodyne. |
| 96 | Sequoia Capital China III | Sequoia China | 2014 | ~$1B | ByteDance (TikTok), Meituan. Exceptional. |
| 97 | GV 2016 | GV (Google Ventures) | 2016 | ~$300M | Various. Uber, Slack follow-ons. |
| 98 | Square Peg Capital Fund III | Square Peg | 2018 | AU$340M | Canva (growth rounds). Australian ecosystem. |
| 99 | Creandum IV | Creandum | 2015 | €175M | Spotify follow-on. Nordic focus. |
| 100 | 500 Global Fund V | 500 Global | 2018 | $140M | Diversified seed. Canva (early). |
Part II: The Patterns — What Fund-Level Data Reveals
The Top 10 Firms by Number of Funds in the Top 100
| Rank | Firm | Funds in Top 100 | Best Fund | Best Fund Performance |
|---|---|---|---|---|
| 1 | Sequoia Capital | 12 | Fund X (1999) | ~8x+ (Google, PayPal) |
| 2 | Benchmark | 5 | Fund IV (2011) | ~11x (Uber, Snap) |
| 3 | Andreessen Horowitz | 6 | Fund I (2009) | ~5x+ |
| 4 | Accel Partners | 4 | Fund IX (2005) | ~20x+ (Facebook) |
| 5 | Kleiner Perkins | 4 | Fund VIII (1999) | ~6x+ (Google) |
| 6 | Union Square Ventures | 4 | Fund 2004 | ~10x+ (Twitter) |
| 7 | Founders Fund | 4 | Fund III (2010) | ~7x+ (SpaceX) |
| 8 | Tiger Global | 3 | PIP XI (2015) | ~3.5x |
| 9 | Blackbird Ventures | 3 | Fund I (2013) | ~92x gross (Canva) |
| 10 | Index Ventures | 3 | Growth III (2014) | ~6x+ (Figma, Roblox) |
The Golden Vintages — When Fund Timing Made Legends
| Vintage Year | Why It Was Special | Best Fund From This Vintage |
|---|---|---|
| 1995-1996 | Pre-dot-com boom. Cheap valuations. | Benchmark I (eBay → 75x fund return) |
| 1999 | Peak dot-com — but Google was in this vintage | Sequoia X (Google, PayPal) |
| 2005 | Post-bust recovery. Web 2.0 dawn. | Accel IX (Facebook → 20x+) |
| 2008-2009 | Financial crisis. Cheapest valuations in a decade. | Sequoia XII (Airbnb, WhatsApp, Dropbox → 9.3x) |
| 2011-2012 | Mobile explosion. Social media monetization. | Benchmark IV (Uber → 11x) |
| 2015 | Cloud maturation. SaaS scaling era. | Multiple strong funds |
| 2019-2020 | Pre-COVID + COVID acceleration. | Founders Fund VII, Thrive VI |
| 2022 | AI revolution begins. Depressed valuations. | Thrive VIII (OpenAI → 126% IRR so far) |
The Fund Size Paradox
One of the clearest patterns in the data: smaller funds generate higher multiples. Every fund in Tier 1 (10x+ TVPI) was under $1 billion in size. Most were under $500 million. The largest — Sequoia XII at $930M — is the exception that proves the rule.
This is mathematical reality. Returning a $100M fund 10x requires creating $1B in value. Returning a $10B fund 10x requires creating $100B in value — essentially building a top-20 global technology company from scratch. It’s not impossible, but it’s astronomically harder.
The mega-fund era (2018-2022) — Tiger Global’s $12.7B fund, Insight’s $20B fund, SoftBank’s $100B Vision Fund — was an experiment in whether throwing massive capital at later-stage companies could generate venture-like returns. The results have been mixed at best, with significant markdowns across the sector.
The “One Company” Phenomenon
The most striking pattern in fund-level data: in many of the greatest funds, a single company drove 70-90% of total returns.
- Accel IX: Facebook was ~90% of fund value
- Benchmark IV: Uber was ~80%+ of fund value
- Blackbird I: Canva was ~95%+ of fund value
- Emergence II: Zoom was ~85%+ of fund value
- Benchmark III: eBay was ~90%+ of fund value
- Thrive VIII: OpenAI is the primary driver
This validates the power law in its most extreme form: venture capital isn’t about building a diversified portfolio. It’s about finding the one company that returns the entire fund many times over, and then not getting in the way.
Part III: The AI Era — A New Paradigm?
The most recent UTIMCO disclosures (November 2025) reveal something unprecedented: AI investments are generating returns at a pace never before seen in venture capital.
Thrive Capital’s 2022 fund posted a 126% IRR — driven primarily by OpenAI. Notable Capital’s 2023 fund swung from -48% IRR to +96% IRR in a single year — driven primarily by Anthropic. As Altimeter’s Meghan Reynolds noted, VC investors’ gross profits on just three LLM companies (OpenAI, Anthropic, and likely xAI or Mistral) currently equate to roughly 70% of all VC profits from the entire previous decade.
This creates a fascinating tension. The best-performing funds of the 2022-2024 vintages are heavily concentrated in AI foundation model companies. If these companies sustain their valuations — if OpenAI’s $300B valuation holds, if Anthropic’s $60B holds — then Thrive VIII and Notable’s 2023 fund could end up among the 10 greatest venture funds ever raised. If AI valuations compress (as they did for SaaS in 2022-2023 and crypto in 2022), these paper gains could evaporate.
The data is both exhilarating and cautionary. Every mega-trend in VC — dot-com in 1999, social media in 2011, crypto in 2021 — has produced both the greatest funds and the most devastating losses. The vintage that created Benchmark I (eBay) also created dozens of dot-com funds that went to zero. The vintage that created Thrive VIII (OpenAI) will inevitably produce AI funds that generate negative returns.
Part IV: The Uncomfortable Truths
Truth #1: Most VC Funds Lose Money
According to Cambridge Associates, approximately 50% of VC funds fail to return invested capital (1.0x TVPI) after fees. The median fund returns roughly 1.5-2.0x. Only the top 25% return 2.5x+. The funds on this list represent the top 1-2% of all funds ever raised.
Truth #2: Past Performance Doesn’t Predict Future Results (Much)
The “persistence” of VC returns — whether a firm’s Fund III performance predicts Fund IV — has declined significantly over the past 20 years. A study by Kauffman Foundation found that only about 25% of top-quartile funds had their next fund also in the top quartile. Sequoia, Benchmark, and Accel are exceptions, not the rule.
Truth #3: Unrealized Returns Are Not Real Returns
Many of the most impressive numbers on this list — Thrive VIII’s 126% IRR, Notable’s 96% IRR, Blackbird I’s 92x gross TVPI — are partially or entirely unrealized. Until cash is distributed to LPs, these are educated guesses. The 2021-2022 vintage taught this lesson painfully: Tiger Global’s Fund XIV showed strong paper returns before marking down by 30%+ as public comps crashed.
Truth #4: The LP Experience Is Very Different from the GP Experience
GPs (general partners — the VCs) earn 2% annual management fees plus 20% carried interest on profits. This means a GP on a $1B fund earns $20M/year in fees before a single dollar is returned to LPs. The GP’s $200M carry on a 3x fund is life-changing wealth. The LP’s 2.4x net return (after fees and carry) might be slightly above what a public market index delivered over the same period.
Truth #5: Access Is the Real Alpha
The most important insight from fund-level data: the gap between the best and worst VC funds is far wider than in any other asset class. The top-decile VC fund might return 10x while the bottom-decile returns 0.3x. In public equities, the gap between top and bottom decile managers is perhaps 3-5% per year. In VC, the gap is 30-50% per year. This means that for LPs, the single most important decision isn’t “should I invest in VC?” — it’s “can I get into the best funds?” Access to Sequoia, Benchmark, or Thrive is worth more than any amount of VC allocation to mediocre funds.
Methodology and Data Sources
This compilation draws from the following public sources:
- UTIMCO (University of Texas Investment Management Company): Public records requests filed by Newcomer and others, covering fund-level IRR data for Sequoia, USV, Thrive, GGV/Notable, IA Ventures, Forerunner, HongShan, Peak XV, and Initialized Capital. Data through November 2025.
- CalPERS (California Public Employees’ Retirement System): Quarterly performance reports for all active PE/VC fund commitments, including net IRR and investment multiples. Data through June 2025.
- PSERS (Pennsylvania Public School Employees’ Retirement System): Fund performance disclosures obtained via public records requests.
- Press reports: TechCrunch, The Information, Newcomer, Bloomberg, WSJ, and others reporting on leaked or disclosed fund performance.
- SEC filings and S-1 prospectuses: Used to estimate ownership stakes and implied fund-level returns at IPO.
- Preqin and Cambridge Associates benchmarks: Used for vintage-year context and industry median comparisons.
- Fund press releases and LP letters: Occasionally disclosed performance metrics from firm blogs and fundraising announcements.
Key Limitations
- Most VC fund returns are private. This list represents publicly known or reasonably estimated data. There are certainly funds that belong on this list but whose performance has never been disclosed.
- IRR is time-dependent and manipulable. Funds can engineer high IRRs through early distributions or capital call timing. TVPI/DPI are more reliable for mature funds.
- Gross vs. net matters enormously. Blackbird I’s ~92x gross TVPI is dramatically higher than its net TVPI to LPs after fees and carry. All figures are noted as gross or net where known.
- Unrealized returns are estimates. Many entries (particularly 2019-2024 vintages) contain significant unrealized value based on last-round markups. These figures should be treated as directional, not definitive.
- Survivorship bias. This list only includes successful or notable funds. For every Benchmark IV, there are hundreds of funds from the same vintage that delivered mediocre or negative returns.
© 2026. All rights reserved. Written for educational purposes only. This is not investment advice. Sources cited within. Fund performance data is based on publicly available disclosures and estimates; actual returns may differ. Not affiliated with any VC firm, pension fund, or endowment mentioned herein.