The average VC fund now holds positions for 12–14 years against a 10-year legal term, and median time-to-meaningful-DPI has stretched past 8 years — up from roughly 5 a decade ago. That's the short answer. The longer answer is more interesting.
The "10-year fund" is the foundational promise of venture capital. LPs commit capital, the fund invests for the first few years, harvests for the rest, and winds down by year 10. Except almost no fund raised in the last decade has actually done that. The structure is intact on paper; the economics underneath it have quietly broken.
VC Fund Lifecycle Problems in 2026: What's Actually Breaking
The core VC fund lifecycle problem in 2026 is duration mismatch. The standard fund term is 10 years with two one-year extensions, so 12 years is the contractual ceiling. But companies now stay private 12+ years before a liquidity event, median time-to-DPI has pushed past year eight, and a growing share of 2012–2015 vintage funds are still holding marquee positions at year 14. The promised lifecycle and the real one no longer line up, and every downstream problem flows from that single gap.
I've been on both sides of this — as a founder whose companies took far longer to exit than anyone modeled, and as an investor with 65+ investments watching marks sit frozen on quarterly reports. The number that matters to an LP isn't TVPI. It's when the wire hits. And that wire is arriving years late.
The Fund Lifecycle on Paper vs. in Reality
Here's how the textbook 10-year fund is supposed to run versus how a typical 2018-vintage fund is actually behaving in 2026:
| Phase | On Paper | Reality in 2026 |
|---|---|---|
| Investment period | Years 1–4 | Years 1–4 (mostly intact) |
| First meaningful DPI | Year 5–6 | Year 8–9 |
| Harvest / exits | Years 5–9 | Years 8–14 |
| Fund fully wound down | Year 10 | Year 13–15 |
| Number of term extensions | 0–1 | 2–4 requested |
| DPI at year 7 | 0.8–1.2x | 0.2–0.5x |
| Companies still private at term | Rare | Common (largest positions) |
The investment period is the only phase that still runs roughly on schedule. Everything after it has slipped by 3–5 years. You can track how vintage-year returns are actually maturing on the VC Performance dashboard.
Why DPI Lags TVPI by So Much Now
The most quoted number in a fund update is TVPI — total value to paid-in — and it's the most misleading. TVPI includes unrealized markups; DPI counts only cash returned. In 2026 the spread between them is the widest it has been in modern venture history. Plenty of 2018–2019 vintage funds report 2.0–2.5x TVPI while sitting at 0.3–0.5x DPI. The paper looks great. The bank account does not.
Three forces drive the gap. First, companies stay private far longer — the median time from founding to IPO crossed 12 years, versus 6–8 years in the 2000s. Second, the IPO window stayed narrow through 2022–2024, so even ready companies waited. Third, late-stage markups inflated TVPI during 2020–2021, and those marks have only partially reset, leaving paper value that hasn't converted to cash. For a deeper benchmark view, the TVPI vs DPI benchmark breakdown shows the divergence by vintage.
12+ yrs
Median founding-to-IPO time, up from 6–8 in the 2000s
2.0–2.5x
Typical TVPI on 2018–2019 vintage funds in 2026
0.3–0.5x
Typical DPI on those same funds — the cash actually returned
How the LP-GP Relationship Breaks Down
When a 10-year fund becomes a 14-year fund, the conflict isn't dramatic — it's a slow erosion of trust across four pressure points. LPs modeled a liquidity schedule that no longer holds, and the GP keeps asking for more time and, often, more fees.
Liquidity planning blown up
LPs (endowments, pensions) modeled distributions by year 7–10 to fund their own obligations. A 4-year delay forces them to sell other assets or stop re-upping.
Fees on frozen capital
Management fees of 0.5–1% on remaining NAV during extension years keep flowing even when no distributions do — LPs pay to wait.
Marks nobody trusts
After repeated extensions, LPs discount GP-reported TVPI heavily. Trust in the marks is the casualty, and that bleeds into the next fundraise.
Re-up decisions on no data
GPs raise Fund IV before Fund II has returned a dollar of DPI, so LPs commit to new funds with no realized proof of the strategy.
The structural problem: GP economics reward AUM and time, while LP economics reward realized cash. A 10-year term was the alignment mechanism that bridged those incentives. Stretch it to 14 years and the alignment frays.
The Workarounds: Extensions, Recycling, and Continuation Funds
The industry has invented three patches for the lifecycle problem. None of them fix the underlying duration mismatch — they redistribute it. Here's what each actually costs LPs:
| Mechanism | What It Does | LP Cost / Risk |
|---|---|---|
| Term extensions | Adds 1–2 years past year 12 with LPAC approval | Continued fees of 0.5–1% on NAV; no liquidity guarantee |
| Recycling | Reinvests early distributions instead of returning them | Delays DPI further; raises effective capital at risk |
| Continuation funds | New vehicle buys aging assets; LPs cash out or roll | Pricing often at a discount to NAV; GP conflict of interest |
| GP-led secondaries | Sells a strip of the portfolio to a secondary buyer | Discounts of 10–30% to NAV common in soft markets |
| Strip sales / partial exits | Sells a portion of a winning position pre-IPO | Caps upside on the fund's best company |
| Dividend recaps | Portfolio company takes on debt to fund a distribution | Adds leverage risk to an unexited company |
Continuation funds are the fastest-growing patch. GP-led secondary volume hit roughly $60B+ in 2024 and kept climbing into 2026, because they let a GP return capital to LPs who want out while holding a company that's still compounding. The catch is the conflict: the same GP sets the price on both sides of the trade. A well-run continuation fund with an independent valuation and a real LP opt-out is a legitimate tool. A rushed one at a steep NAV discount is a wealth transfer dressed as liquidity.
What LPs Should Demand Before Agreeing to an Extension
Extensions aren't inherently bad — a fund holding a company about to IPO at year 13 should absolutely keep holding it. The question is whether the remaining assets justify the wait and the fees. Here's the line I'd draw:
Say Yes When
- ✓ A clear roadmap to DPI within 2–3 years
- ✓ Remaining value concentrated in 1–3 credible names
- ✓ Fees stepped down or waived during extension
- ✓ Independent third-party valuation of the marks
Push Back When
- ✕ Blanket extension with no liquidity timeline
- ✕ Full fees continuing on stale NAV
- ✕ Value spread thinly across zombie positions
- ✕ Marks unchanged for 8+ consecutive quarters
Is the 10-Year Fund Structure Dead?
Not dead — mispriced. The 10-year term made sense when companies exited in 6–8 years. With median founding-to-IPO past 12 years, the honest structure is a 12-year fund with explicit extension economics negotiated up front, not bolted on at year 11 under duress. Some emerging managers are already raising on 12–15 year terms with built-in continuation rights, and evergreen and rolling-fund structures sidestep the fixed-term problem entirely.
The funds that will keep LP trust are the ones that stop pretending. Tell LPs the real duration, price the extension fees up front, and report DPI as the headline number instead of TVPI. The ones that keep selling a 10-year story they can't deliver will find their next fund a lot harder to raise. You can compare how different vintages and fund types are tracking on realized versus paper returns on the VC & PE Performance dashboard.
The 10-year fund isn't a contract anymore. It's a marketing slogan.
The funds that survive the next cycle will be the ones that price the real 14-year duration honestly — and report DPI, not TVPI, as the number that counts.
Track realized vs. paper venture returns on the VC Performance Dashboard at Value Add VC. Originally published in the Trace Cohen newsletter.