How LPs Choose VC Funds: Why 79% of Re-Ups Get Rejected and What Wins a Commitment

79% of LPs declined to re-up with at least one existing VC manager in the past 12 months. That's the short answer to why fundraising feels brutal right now even for funds with decent numbers. The longer answer is that LPs have quietly rewritten the scorecard, and most GPs are still pitching to the old one.

Coller Capital's Winter 2024-25 Global Private Capital Barometer put a hard number on something every fund manager raising in 2026 already feels in the room: LPs are pickier, slower, and far more willing to walk away from an existing relationship than they were three years ago. At the same time, PitchBook data shows capital concentrating harder than ever into a small set of firms — 12 US venture firms captured more than half of all VC dollars raised in the first half of 2025. Those two facts together are the actual LP decision framework. This post breaks down what it looks like in practice.

How LPs Choose VC Funds to Back

LPs choose VC funds by weighting realized returns (DPI) and re-up track record above paper markups, then layering in team stability, sourcing discipline, and how concentrated the LP wants its own book to be. Coller Capital's data shows 79% of LPs declined at least one re-up in the past year — 42% citing performance, 29% citing their own capital constraints, and 16% citing a strategy or team change at the GP.

That's a meaningfully different posture than five years ago, when a fund's second or third vehicle was close to an automatic re-up if the first fund wasn't underwater. Today, every fund in market — regardless of vintage number — is effectively being evaluated as if it were a first-time close. That shift alone explains why fundraising cycles have stretched even for managers with respectable numbers on paper.

It also explains why the 21% of LPs who did re-up with every existing manager relationship in Coller's Winter 2024-25 survey window look different from the LPs who didn't: they tend to be allocators who deliberately concentrated their book into fewer, higher-conviction managers years earlier, rather than spreading commitments thin across a large roster. The lesson for GPs is blunt — the LPs most likely to keep backing you across cycles are the ones you built the deepest relationship with early, not the ones you added most recently to round out a fund.

DPI Is Now the Deciding Metric, Not IRR

For most of the 2010s, IRR and TVPI drove LP decisions — a fund could point to a fast markup on a hot deal and get credit for it in a re-up conversation. That era is over. Bain & Company's Global Private Equity Report 2026, surveying roughly 300 institutional LPs, found DPI has risen to tie with MOIC as the second most heavily weighted metric behind IRR itself. The reason is simple: distributions have dried up. McKinsey's Global Private Markets Report 2026 found distributions ran at just 6% of AUM in the 12 months ending June 2025, compared to a 14% average across 2015-2024 and 16% across 2015-2019.

Bain's own data shows distributions as a share of NAV have held below 15% for four consecutive years. When cash isn't coming back, LPs stop rewarding paper gains and start demanding proof — which is exactly why a fund's realized DPI now carries more weight in a re-up decision than its unrealized TVPI. For a deeper look at how these metrics interact, see our post on how VC fund performance is actually measured.

The LP Decision Framework, by the Numbers

Pulling together the re-up data, the fundraising concentration data, and the performance benchmarks, the actual framework institutional LPs are applying in 2026 looks like the table below. It's less a single checklist than a set of filters a fund has to clear in sequence — and the filters have gotten materially harder to pass since 2020.

Figures are 2020-2026 estimates blended from Coller Capital's Global Private Capital Barometer, PitchBook-NVCA Venture Monitor, Cambridge Associates benchmark commentary, and McKinsey's Global Private Markets Report 2026. TVPI ranges reflect commonly cited benchmark spreads for mature vintages, not a single named report.

Why Capital Concentration Makes the LP Decision Framework Harsher

The re-up data explains why existing managers are getting scrutinized harder. The concentration data explains why new managers have it even worse. PitchBook found that in full-year 2024, the top 9 US venture firms alone raised $35 billion — roughly half of all US VC fund capital raised that year — while the top 30 firms captured about 75% of the total. In the first half of 2025, that concentration held: just 12 firms raised more than 50% of all capital, and the top 30 captured 74%.

Founders Fund alone raised $4.6 billion in 2025 — more than double what all 44 first-time fund managers collectively raised from LPs that same year, according to PitchBook and Axios Pro coverage. That's the sharpest illustration of how the LP decision framework actually functions in practice: when capital is scarce and re-ups are being rejected at a 79% clip, LPs default to writing bigger checks to fewer, larger, already-proven names rather than spreading smaller checks across a wider set of emerging managers. Anyone raising a first or second fund right now is competing directly against that gravitational pull. Our VC & PE performance dashboard tracks how these dynamics show up across fund vintages.

The LP Decision Framework in Practice: A Founder's and GP's Guide

If you're a GP raising in 2026, the practical version of the LP decision framework is: lead with DPI, not TVPI. A fund that has returned real cash — even at a modest multiple — now beats a fund sitting on unrealized markups at a higher multiple, because LPs have been burned by paper gains that never converted to distributions during the 2021-2023 markdown cycle. Second, expect diligence to take longer and go deeper: family offices typically move in 4-8 weeks, funds-of-funds in 8-16 weeks, and pensions or endowments in 12-24 weeks, with reference calls to portfolio-company founders now a standard, not optional, step.

Third, be honest about where you sit in the concentration curve. If you're outside the top 30 US venture firms by AUM, you're fundraising against a market where those firms alone captured three-quarters of all committed capital in 2024 and H1 2025 combined. That doesn't mean smaller or emerging managers can't raise — Founders Fund's $4.6 billion outlier year proves scale isn't destiny — but it does mean the pitch has to be sharper: a differentiated thesis, a demonstrable sourcing edge, and early proof points on DPI, not just IRR on paper. Fund managers can stress-test their own numbers against current benchmarks using our fund benchmarking tool.

For founders raising a company, the same framework matters indirectly — it determines which VCs are actually flush with fresh capital and which are stretched thin waiting on their own LP re-ups. A fund that just closed on the back of strong DPI has dry powder and conviction; a fund mid-raise with a shaky re-up rate may be slower to lead, slower to follow on, and more conservative on terms. Asking a prospective lead investor when they last closed a fund, and how much of it came from returning LPs versus new ones, is a fair diligence question founders rarely think to ask but should.