The headline: fundraising is fine, distribution is not

US venture firms raised $72.4 billion across 405 funds in the first half of 2026, according to the Q2 2026 PitchBook-NVCA Venture Monitor as reported by SiliconANGLE. All of 2025 came in at $74.9 billion. In other words, six months of 2026 nearly matched the full prior year.

If you only read that total, you would conclude that capital is easy to find. The breakdown says otherwise. Funds of $1 billion or more accounted for $49.5 billion of the H1 total, roughly two-thirds of everything raised. Three firms alone drew $34.8 billion, or 48% of H1 fundraising: Andreessen Horowitz closed seven funds worth $14.2 billion, Founders Fund raised $10.6 billion across two funds, and Thrive Capital raised $10 billion across two.

Foley & Lardner's read of the same data puts the large-fund share at 68.3%, up from 36.1% in 2025, and says twelve firms accounted for about three-quarters of all dollars raised. The exact count of firms depends on how the data is cut, but the direction is unambiguous: the market is splitting into a few very large pools and a long tail of small ones.

The long tail is crowded and thinly funded

Foley reports that funds under $50 million made up 67.7% of closed funds, a decade high, yet raised only 4% of the money. First-time fund closings hit a decade low. Carta's data on fund performance showed a similar shift last year, with at least 40% of closed funds in 2024 and 2025 sitting in the $1 million to $10 million range, up from 25% in 2020.

For emerging managers, this is the practical reality. Plenty of people are able to close a small vehicle, often from angels and a few family offices. Very few are able to graduate to a $50 million or $100 million fund, because institutional limited partners (LPs) are concentrating their commitments with managers that already have track records and access to the largest AI rounds.

Why LPs are behaving this way

Two forces push institutions toward fewer, bigger relationships.

Deal concentration. US companies raised $412.7 billion in H1 2026, 29% more than all of 2025 and 15% above the 2021 record, per Axios. More than 81% of those dollars went to rounds of $100 million or more. The Houlihan Capital summary of the Venture Monitor puts AI at about 86% of deal value. Investing in that environment needs large checks and reserves, which favors large funds.

Thin liquidity. Foley's summary of the PitchBook data says cash distributed has run at 11.7% to 12.8% of asset value, against a long-run average of 18.7%. The one-year US venture return was 17.1%, but the three-year return was only 5.7%. LPs that have not been getting cash back are being selective about where new money goes, and many prefer to re-up with managers they already know.

Exits are improving, but unevenly. SpaceX raised $75 billion in its IPO at a $1.7 trillion offering value, and PitchBook called it the largest IPO ever. Houlihan notes that Q2 exit value would look far smaller without that single transaction. Foley adds that venture-backed IPOs have numbered fewer than 50 in any year since 2022. One mega-listing does not mean broad distributions to LPs.

What it means for valuations and terms

The same concentration shows up in pricing. Carta's Q1 2026 data put the median Series A valuation at about $300 million for AI foundational model companies and about $55 million for non-AI companies. Down rounds were 11.4% of rounds in Q1, back in line with 2019 and 2020 and well under the 22% peak in 2023.

That is the picture for companies that can raise. For funds, Foley describes a market where LP-friendly terms are getting harder to hold: thinner side letters, fee and recycling terms moving through with less negotiation, and more structure on late-stage term sheets such as preferences, ratchets and participation rights. The article is qualitative on this, with no numeric terms, so treat it as a signal rather than a measurement.

Playbook notes for the people raising

  • Founders: Which investors have fresh capital matters as much as the firm's brand. Funds that closed in the last 12 months have reserves for follow-ons; older vehicles may not. Ask how much of the fund is still uninvested and how much is reserved for existing companies.
  • Emerging fund managers: Expect a longer, smaller first close. Carta's fund data shows how common $1 million to $10 million vehicles have become, which means your differentiation has to be a specific edge in sourcing or sector, not a general thesis.
  • CFOs of later-stage companies: The late-stage structure Foley describes means headline valuations can hide protections that matter in a downside. Model the exit waterfall, not only the post-money.
  • Investors tracking the asset class: Watch distributions rather than markups. Foley's note that fund asset values grew 21.6% in 2025, largely from AI markups, is a reminder that paper gains have not yet become cash.

What to watch next

The Q3 2026 PitchBook-NVCA Venture Monitor should show whether H1's mega-fund concentration was a one-time effect of a handful of closes or the new pattern. Three things matter most: whether full-year fundraising clearly passes 2025, whether first-time fund closings recover from their decade low, and whether the confidential IPO filings that PitchBook says Anthropic and OpenAI have made turn into listings that return cash to LPs.

Until distributions pick up, the likely outcome is more of the same: abundant capital at the top, scarce capital at the bottom, and the middle of the market doing the hard work of proving it deserves an allocation.

This article is information and commentary, not investment advice.

Sources