VC fund performance is measured with three numbers — net IRR, TVPI, and DPI — and in 2026 the honest read on all three is uncomfortable. The median 2021-vintage fund sits at 1.02x TVPI, a 1.4% net IRR, and essentially zero cash returned to LPs. Meanwhile a genuinely top-quartile mature fund still clears 2.6x-3.6x TVPI and 21%-29% net IRR.
That spread — between the paper story and the cash story, and between the median and the top quartile — is the whole game. This guide walks through how each metric works, what the current median and top-quartile benchmarks are by vintage year using Carta's Q1 2026 dataset of 2,500+ funds plus Cambridge Associates and PitchBook data, and why LPs quietly replaced TVPI with DPI as the metric that decides re-ups. If you want to check a specific fund against these numbers, our free Fund Benchmarking tool does the quartile comparison for you.
Sources: Carta VC Fund Performance Q1 2026 (2,500+ funds on Carta Fund Administration); PitchBook Benchmarks; Cambridge Associates US Venture Capital Index.
How Is VC Fund Performance Measured?
Every venture fund is judged on three core metrics, and each answers a different question:
- Net IRR (internal rate of return) — the annualized return LPs earn after fees and carry, sensitive to timing. It answers: how fast is my money compounding?
- TVPI (total value to paid-in capital) — distributions plus the current marked value of unrealized holdings, divided by capital paid in. It answers: what is everything worth on paper today?
- DPI (distributions to paid-in capital) — cash actually wired back to LPs, divided by capital paid in. It answers: how much real money have I gotten back?
A useful companion is RVPI (residual value to paid-in) — TVPI minus DPI, i.e. the part of the multiple that is still just a mark. Early in a fund's life, TVPI and IRR are all you have. Late in a fund's life, DPI is all that matters. The mistake most people make is comparing funds of different ages on the same metric — a 3-year-old fund at 1.3x TVPI and an 8-year-old fund at 1.3x TVPI are completely different stories. That's why every serious benchmark, from Cambridge Associates to PitchBook to Carta, cuts the data by vintage year: the year the fund made its first capital call, which cohorts funds against peers that faced the same entry prices and the same exit market.
VC Fund Performance Benchmarks by Vintage Year (2026)
Here is the current state of play by vintage, blending Carta's Q1 2026 medians for recent vintages with PitchBook and Cambridge Associates quartile data for mature ones. Medians are net of fees and carry.
| Vintage | Fund Age | Median TVPI | Median DPI | Median Net IRR | Top-Quartile Marker |
|---|---|---|---|---|---|
| 2016 | ~10 yrs | ~1.8x | well below 1x for most | ~12-15% | ~3.6x TVPI / ~29% IRR |
| 2017 | ~9 yrs | ~1.7x | 0.31x | ~10-13% | ~3x+ TVPI |
| 2018 | ~8 yrs | ~1.6x | 0.15x | ~9-12% | ~3x TVPI |
| 2019 | ~7 yrs | ~1.4x | barely above zero | ~6-9% | ~2.8x TVPI |
| 2020 | ~6 yrs | ~1.3x | near zero (42% have any DPI) | ~4-7% | ~2.6x TVPI / ~21% IRR |
| 2021 | ~5 yrs | 1.02x | near zero (90th pct: 0.16x) | 1.4% | 1.4-1.6x TVPI |
| 2022 | ~4 yrs | 1.02x | near zero | 0.7% | ~1.3-1.5x TVPI |
| 2023-24 | 1-3 yrs | ~1.0-1.1x | n/a (too early) | n/a (J-curve) | 90th pct IRR >20% |
Medians for 2017-2024 vintages from Carta's quarterly VC Fund Performance reports (Q1 2025 through Q1 2026, 2,500+ funds); mature-vintage medians and top-quartile markers are approximations blended from PitchBook Benchmarks and Cambridge Associates quartile data as of mid-2026. Treat mature-vintage figures as directional — exact quartile cuts vary by provider and reporting quarter.
Two things jump out. First, the 2021-2022 cohort — the largest vintage by capital raised in venture history — is barely above water on paper and has returned almost nothing in cash. Second, dispersion is enormous: Carta's data shows the 90th-percentile net IRR above 20% for nearly every vintage from 2017 through 2024 (2021 being the lone exception), while Cambridge Associates data puts the top-to-bottom quartile spread at more than 30 percentage points of IRR in most vintages — wider than any other private-market strategy. In venture, which fund you picked matters vastly more than whether you were in venture.
What Do Median vs Top-Quartile Returns Actually Look Like?
The classic top-quartile end-state — the bar LPs use when underwriting a new commitment — is roughly 3.0x+ net TVPI, 20-25%+ net IRR, and DPI north of 1x by the back half of the fund's life. The long-run median fund, by contrast, delivers about 1.5x-1.8x TVPI and 12-15% net IRR — real but hardly compensation for a decade of illiquidity when public equities were compounding comparably. The Cambridge Associates US Venture Capital Index, the institutional yardstick, gained 6.4% in the first half of 2025 as the recovery took hold, and Carta's Q1 2026 report showed median TVPI rising for nearly every vintage from 2017 through 2024 — the broadest quarterly improvement since the correction began. The trend is finally up; the base it's rising from is low. We wrote a deeper cut on the top decile in what top-quartile IRR, TVPI, and DPI actually look like.
Why Does DPI Now Matter More Than TVPI?
Because for three straight years, the paper didn't turn into cash. When the IPO window shut in 2022, average hold periods stretched from roughly 4.5 years to over 7, and annual distributions have been stuck at about 14-15% of NAV since late 2022 — roughly half the historical norm. The result is what LPs call the DPI drought: the 2021 vintage as a cohort has returned on the order of 0.08x — eight dollars back for every hundred committed, five years in — with hundreds of billions in unreturned capital sitting in aging portfolios.
Even the vintages that should be harvesting are behind. Per Carta's Q1 2026 data, the 2017 vintage's median DPI is just 0.31x and 2018's is 0.15x — nine and eight years in — with fewer than 20% of those funds past 1x. The 2019 and 2020 vintages' median DPI is barely above zero, and fewer than half of those funds have returned any capital at all. Historically, the average VC fund reaches only ~0.7x DPI by year 8, versus ~1.3x for the average PE fund. When an LP can't get cash out, they can't recycle it into new commitments — which is why the drought flowed straight through to the brutal fundraising market emerging managers have faced since 2023.
So the hierarchy inverted. A decade ago GPs raised on IRR and TVPI; today the first question in every LP meeting is "what's your DPI, and what's the path to more?" There are green shoots — Carta's Q3 2025 report showed 42% of 2020-vintage funds with a first distribution (up from 28% a year earlier) and 25% of 2021 funds distributing (up from 14%), and if the 2026 IPO reopening and secondary-market pipeline hold, the coming cycle could be one of the largest LP payout waves in venture history. But until the cash shows up, DPI trajectory — not TVPI level — is what gets a re-up. We track the whole story live on our DPI Crisis dashboard.
Median 2021-vintage fund: 1.02x TVPI, 1.4% net IRR, near-zero DPI. Top-quartile 2016-vintage fund: ~3.6x TVPI, ~29% net IRR.
In venture, the average fund barely works. The top quartile is the entire asset class.
How Do You Benchmark a Specific VC Fund?
The method is simple even if the data access usually isn't: take the fund's net TVPI, DPI, and IRR from its LP report, and compare each against the quartile breaks for the same vintage year — never against the all-vintages average, and never against a fund three years older. A 2021 fund at 1.3x TVPI is comfortably above its cohort's median (1.02x) and pushing toward its top quartile (1.4x-1.6x), even though 1.3x would be a disappointing number for a 2016 fund. Then check the DPI trajectory: is the fund's distribution line moving, or is the multiple all residual value?
The institutional sources — Cambridge Associates, PitchBook Benchmarks, Preqin — sit behind subscriptions, while Carta publishes its quarterly medians free. To make this practical, we built a free VC Fund Benchmarking tool that puts a fund's TVPI, DPI, and IRR against same-vintage top-quartile and median data in one view — it's the fastest way to see where a fund actually stands. If you're a GP preparing for LP meetings, our guide on fund benchmarking for emerging managers covers how to present these numbers, and if you're hunting for raw data, here's where to find fund performance data without a Bloomberg terminal.
The Bottom Line
VC fund performance in 2026 is a tale of three numbers moving at three speeds. TVPI is recovering — medians rose across nearly every vintage in Q1 2026. IRR is crawling back into positive territory for the peak-era vintages. And DPI, the only number that ever mattered in the end, is still the bottleneck: most funds raised since 2019 have returned little or nothing in cash. Benchmark any fund against its own vintage, weight the cash over the paper, and remember the 30-point quartile spread — in this asset class, manager selection isn't part of the strategy, it is the strategy. Run your own numbers on the Fund Benchmarking tool.
Benchmark any VC fund's TVPI, DPI, and IRR against same-vintage quartiles with the free Fund Benchmarking tool at Value Add VC. Reach out at t@nyvp.com or @Trace_Cohen.
Get VC data most people never see
— 100% free
Weekly benchmarks, valuations, and fund data. Join 5,000+ investors. No spam.