VenCap International

VenCap International plc is a UK fund-of-funds manager founded in 1987 and investing in venture capital since 1989 — nearly four decades and multiple technology cycles, from the PC era through the dot-com boom and bust to the cloud, mobile and now AI waves. It runs roughly $2.2bn and has committed to more than 100 underlying funds across its history.

Its model is deliberately narrow. Rather than chase the long tail of emerging managers, VenCap concentrates on a curated set of roughly 100 "Core Managers" — the top-tier venture firms it believes demonstrate genuine return persistence — and re-ups with them fund after fund. The discipline is in the restraint: no spray-and-pray across first-time funds, no style drift, just sustained access to a small group of GPs that have repeatedly caught the winners. For an allocator, that access is the scarce resource — most of these firms are closed to new investors, and a multi-decade LP relationship is what opens the door.

VenCap's intellectual core is an empirical study of its own deal flow — 11,350 startups across 259 funds from 1986 to 2018 — and the finding is stark:

The conclusion follows directly: a great venture fund almost always rests on catching one of the rare, exponential winners — and those are hard to reach except from inside the small number of firms that consistently land them. Venture is less a diversification game than an access-to-outliers game.

The current AI cycle is the thesis playing out in real time: a company like Anysphere (Cursor) moving from a ~$2.5bn round to ~$29bn in roughly a year is the kind of exponential move the strategy is built to capture — and only the top managers had the access to write those early cheques.

Across the fifteen VenCap funds with realised performance, the record is strong and, notably, multi-cycle:

Three things stand out. First, the best vintages are exceptional for a fund-of-funds — the 2014 and 2015 vehicles returned 3.8× and 4.4×. Second, the mature funds have actually returned cash (DPI around or above 100%), not just paper marks. Third, the record is honest about its weak spots: the 2000-vintage funds, raised into the dot-com peak, did roughly nothing — even a strong manager is hostage to vintage timing, and pacing across years matters more than picking one entry point.

This corroborates VenCap's own headline — roughly 3.4× on its mature "Core Manager" portfolio — as the core-weighted top end of a record whose broad median sits near 2.4×. As an external sanity check, the British Business Bank has found only about 16% of UK venture funds ever reach 3×; VenCap's mature funds clustering at or above that mark is the signal that matters.

A fund-of-funds is only as good as the funds inside it. So the Core Managers can be set against the entire venture universe — 3,760 funds with realised IRRs (median 10.5%, upper-quartile 21.8%, top-decile 38.2%) — ranking all 226 managers with five or more scored funds by their top-quartile hit rate, the fairest cross-vintage measure of how often a firm is elite.

The top of the roster is genuinely strong: Index ranks 2nd and Sequoia 10th of 226, both top-5% managers, with Kleiner in the top 15% (albeit on legacy-era funds). Below that the picture is uneven — several names (Khosla, Lightspeed, Spark) sit around the universe average on realised persistence, and Peak XV's recent funds land in the bottom third. And the two firms most central to the current AI cycle — Andreessen Horowitz and Thrive — have no realised track record in the data, because a16z discloses little publicly and Thrive's funds are young; their quality rests on reputation and current marks, not yet on harvested returns.

The evidence, in short, supports the quality of VenCap's best relationships without supporting the idea that every Core Manager is uniformly elite. A handful are; the rest range from very good to merely average.

VenCap 18 is the latest vintage of the flagship programme (2024 vintage, ~$350m target), built to capture this cycle's outliers through the Core Managers. Its construction is disciplined: at least 75% to Core Managers, a three-year deployment window across the 2025–2027 fund vintages, a balance of early and growth exposure (growth shortens duration and lifts IRR), and 10–20% in secondaries to soften the J-curve.

What makes this vintage distinctive is its early, indirect exposure — via the underlying GPs — to a cluster of the defining companies of the AI era, several entered at prior-round valuations below where they trade today. Per VenCap's reporting, the look-through portfolio includes positions such as OpenAI, Databricks, Anysphere (Cursor), SpaceX, Wiz, Anduril and Stripe, accessed largely through Andreessen Horowitz and Thrive (together roughly half the fund), alongside Kleiner Perkins, Khosla, Accel and Peak XV.

The case for a manager like this comes with real caveats. The main ones:

VenCap is a four-decade, multi-cycle allocator whose record holds up to scrutiny — mature funds clustering at 2.4–4.4×, with cash genuinely returned rather than only marked. Its philosophy is coherent and broadly data-supported: venture is an access-to-outliers game, and its edge is sustained access to the firms that catch them. Its strongest relationships — Index, Sequoia — rank among the best managers in the entire venture universe.