
VC fund performance metrics measure how much value a venture fund has created for its limited partners (LPs) and how fast. The core set is TVPI, DPI, RVPI, IRR, MOIC and PME. TVPI and IRR show paper progress, DPI shows cash actually returned, and PME shows whether the fund beat a public index. In our view, no single metric is enough on its own.
Definition: VC fund performance metrics are standardized ratios and return calculations that compare what LPs have paid into a fund with what they have received and what the fund still holds, over time and relative to public markets.
Every fund deck has a flattering number on it. The job is spotting what it leaves out.
The multiples: TVPI, DPI, and RVPI
The three multiples use standard formulas:
- DPI (distributions to paid-in capital) = cumulative distributions divided by paid-in capital. This is realized return: cash or stock actually sent to LPs.
- RVPI (residual value to paid-in capital) = residual value (current NAV) divided by paid-in capital. This is unrealized return: what the remaining portfolio is marked at.
- TVPI (total value to paid-in capital) = distributions plus residual value, divided by paid-in capital. So TVPI = DPI + RVPI.
"Paid-in" means capital actually called, not the full commitment. A fund that has called 60 percent of commitments measures against that 60 percent.
Worked example
An illustrative case: a $50M fund has called $40M. It has distributed $20M from two exits, and the remaining portfolio is marked at $60M.
| Metric | Calculation | Result |
|---|---|---|
| DPI | $20M / $40M | 0.5x |
| RVPI | $60M / $40M | 1.5x |
| TVPI | $80M / $40M | 2.0x |
A 2.0x TVPI sounds strong. But three quarters of it is still marks. If the largest holding is revalued downward, TVPI can drop quickly while DPI stays fixed. DPI only moves when a liquidity event turns a mark into cash, which is why a common LP saying holds that DPI is the one multiple you can spend.
IRR: the time dimension
IRR (internal rate of return) is the annualized discount rate that makes the present value of all contributions equal the present value of all distributions plus current NAV. It rewards speed: returning 2x in three years produces a far higher IRR than returning 2x in ten. For the formula and spreadsheet steps, see internal rate of return (IRR) explained.
IRR has two well-known weaknesses in venture:
- It can be managed. Subscription credit lines delay capital calls, which shortens the measured holding period and lifts IRR without changing the multiple. ILPA's Principles 3.0 say LPs should receive IRR and TVPI or MOIC figures both with and without the effect of such facilities.
- It's unstable early. In the first few years, a single markup can produce an eye-catching IRR on a small base of paid-in capital.
A huge IRR in year two is a headline. It isn't a result yet.
MOIC and gross versus net returns
MOIC (multiple on invested capital) is usually measured at the deal or portfolio level: the value of an investment divided by the amount invested in it. AngelList notes it is calculated before fees, expenses and carried interest. The exit multiple guide covers deal-level multiples and how they are underwritten.
Check which version you're looking at:
- Gross returns measure the portfolio before fund costs.
- Net returns measure what LPs actually receive after management fees, expenses, and carried interest.
The gap can be large. Cambridge Associates' paper on investment-level benchmarks, which covers private equity funds rather than venture, estimates that fund-level fees cost investors about 616 basis points on average, or roughly 0.2x of MOIC, over a fund's life. We'd choose between funds on net numbers. For how carry is calculated, see venture capital carried interest.
ILPA's new Performance Template, released in January 2025, standardizes this: it sets out how GPs report IRR and TVPI or MOIC, gross and net, with and without the impact of subscription facilities. ILPA's stated timeline is for the new templates to be implemented beginning in Q1 2026.
PME and KS-PME: did the fund beat the index?
A public market equivalent answers a simple question: what would LPs have earned if they had put the same cash, on the same dates, into a public index instead?
Carta's PME guide describes three common methods. Long-Nickels PME (LN-PME) and PME+ convert the fund's cash flows into index trades and report a comparable IRR. The Kaplan-Schoar PME (KS-PME) produces a ratio instead: grow each contribution and distribution at the index's return to the measurement date, then divide the future value of distributions plus NAV by the future value of contributions. Carta notes a KS-PME above 1.0 means the fund beat the benchmark. Cambridge Associates uses its own modified PME (mPME), which buys and sells index shares on the fund's cash flow schedule.
KS-PME worked example
In this illustrative example, a fund calls $10M at year 0 and $10M at year 2, distributes $15M at year 5, and holds $20M of NAV at year 6. The public index stands at 100, 110, 140 and 150 on those dates.
- Future value of contributions: $10M x 150/100 + $10M x 150/110 = $15.0M + $13.6M = $28.6M.
- Future value of distributions plus NAV: $15M x 150/140 + $20M = $16.1M + $20.0M = $36.1M.
- KS-PME: $36.1M / $28.6M = 1.26.
The fund's TVPI is 1.75x ($35M over $20M) and its IRR is about 12.9 percent. A KS-PME of 1.26 means LPs ended with about 26 percent more value than the index would have delivered.
PME matters more right now, in our view. Cambridge Associates' commentary for calendar year 2025 reported a 21.1 percent return for its US Venture Capital Index (a pooled IRR, net of fees, expenses and carry), its best year since 2021, and found that venture outperformed public indexes over shorter periods but has been less consistent across the periods shown, particularly against the tech-heavy Nasdaq.
VC fund performance benchmarks for 2026
Benchmarks are most meaningful by vintage year, because a 2023 fund and a 2016 fund are at different points in their lifecycle. Carta's VC Fund Performance report for Q1 2026 (published June 2026, covering 2,775 funds) is the latest in the series:
- Top-decile IRR. For every vintage from 2017 through 2024 except 2021, the 90th percentile net IRR is above 20 percent, while the 75th percentile is not above 15.5 percent in any of those vintages.
- Cash returned. In the 2017 and 2018 vintages, fewer than 20 percent of funds have reached 1x DPI. For 2019 and 2020 funds, median DPI is barely above zero and fewer than half of funds have returned any capital.
- Trend. Median net TVPI rose for nearly every recent vintage in Q1 2026, and has climbed steadily over the past six quarters for every vintage from 2017 through 2024.
Cash is starting to flow back. Cambridge Associates reports that US venture managers called $61 billion from LPs in 2025 and returned $42 billion.
The shape of returns explains the wide spread. Andreessen Horowitz's 2015 analysis of Horsley Bridge data found that about 6 percent of investments, representing 4.5 percent of dollars, produced roughly 60 percent of total returns. Funds that land one of those outliers often move to the top quartile; funds that miss struggle to recover with singles. That's the power law we keep coming back to, and venture capital portfolio strategy covers how GPs build around it.
"But cash is all that counts"
After years of marks that rarely turned into money, many LPs now say this, and it's easy to see why. DPI can't be revised by a friendly valuation.
But Judge a young fund on DPI alone and nearly every recent vintage fails. A near-zero median DPI for 2019 and 2020 funds says more about exit markets and fund age than about manager skill. Our read: weight DPI more heavily as a fund ages, and in the early years look hard at the quality of the marks behind TVPI. Who set them, when, and on what round?
How LPs might read VC fund performance metrics: a 7-step checklist
- Confirm the vintage and compare to the same vintage benchmark.
- Separate DPI from RVPI and ask what share of TVPI sits in the top two holdings.
- Check net, not gross, and confirm carry accruals are deducted.
- Ask for IRR and TVPI with and without any subscription line.
- Review the marks: when was each top holding last priced, and by whom? The venture capital fund accounting guide explains how those NAV marks are set.
- Compare to a PME such as KS-PME for the same cash flows.
- Track the trend quarter over quarter rather than reacting to one report.
Reading these reports alongside a working team builds judgment faster, in our view. 1752vc's Emerging Angels is an 8-week live program for accredited investors who are new to angel investing, giving them a seat at the table in a working fund's process, with live diligence calls, deal reviews, monthly Investment Circles, and a private community. More fund guides live in the venture capital section.
The bottom line
Each metric answers a different question: paper value, cash home, speed, and whether the fund beat the easy alternative. Read them together, net of fees, against the right vintage, and be skeptical of whichever one the manager puts first.
Marks tell you what a fund might be worth.
Distributions tell you what it was worth to you.
Key takeaways
- VC fund performance metrics combine multiples (TVPI, DPI, RVPI), time-weighted returns (IRR), and a public comparison (PME).
- TVPI equals DPI plus RVPI, so a high TVPI with low DPI is mostly unrealized value.
- IRR rewards speed and can be lifted by subscription credit lines, so ILPA recommends asking for it with and without the facility.
- A KS-PME above 1.0 means the fund beat the public index on the same cash flows.
- Carta's Q1 2026 data show top-decile venture funds above 20 percent net IRR in most 2017 to 2024 vintages, but fewer than 20 percent of 2017 and 2018 funds have returned 1x in cash.
- As a rule of thumb, benchmark net returns against the same vintage year.
Frequently asked questions
The main VC fund performance metrics are TVPI, DPI, RVPI, IRR, MOIC and PME. The multiples show total, realized and unrealized value per dollar paid in, IRR adds time, and PME compares the fund with a public index. Many LPs read them together because each can mislead alone.
A net TVPI above 1.0x means LPs are ahead after fees and carry, but "good" depends on the fund's age. Young funds often sit near or below 1x because fees are paid before companies are marked up. In our view, a fairer test is how the fund ranks against the benchmark for its own vintage year, such as Carta's or Cambridge Associates' data.
DPI counts only distributions LPs have actually received, divided by paid-in capital. TVPI adds the current value of unsold holdings. The difference is RVPI, the unrealized part of the return, which can still rise or fall before an exit.
DPI is cash in hand, so a new valuation cannot revise it. With longer holding periods, many funds show healthy TVPI but little DPI: Carta's Q1 2026 data show median DPI barely above zero for 2019 and 2020 vintages. LPs need distributions to pay for new fund commitments, so DPI appears to weigh more and more in re-up decisions.
Targets vary by strategy, and early-stage managers aim well above public equity returns to justify illiquidity. Results are lower for most funds: Carta's Q1 2026 data show 90th percentile net IRR above 20 percent for most 2017 to 2024 vintages, while the 75th percentile is no higher than 15.5 percent.
KS-PME, the Kaplan-Schoar public market equivalent, divides the index-adjusted value of a fund's distributions plus NAV by the index-adjusted value of its contributions. A result above 1.0 means the fund outperformed the chosen public index on the same cash flows, and below 1.0 means an index fund would have done better.
Sources
- Carta: VC Fund Performance, Q1 2026
- Cambridge Associates: US PE/VC Benchmark Commentary, Calendar Year 2025
- Carta: Public Market Equivalent (PME), From Calculation to LP Reporting
- AngelList: What to Know About TVPI
- ILPA: Principles 3.0, Fostering Transparency, Governance and Alignment of Interests (PDF)
- ILPA: Updated Reporting Template and New Performance Template Released
- Cambridge Associates: Investment-Level Benchmarks for Private Investment Performance Measurement (private equity dataset)
- Andreessen Horowitz: Performance Data and the Babe Ruth Effect in Venture Capital
Disclaimer: This guide is for general education only and is not legal, tax or investment advice. Laws, market data and program terms change, so it may not reflect the latest developments or fit your situation. Treat it as a starting point, not a source of truth, and talk to a qualified lawyer, accountant or financial adviser before you make decisions.


