Omri Casspi's $250m Swish fund shows how solo investors now win institutional money

Swish Ventures, the firm founded in early 2025 by Omri Casspi, the first Israeli to play in the NBA, said on Thursday, September 3, that it had closed its third fund at $250m. The firm says the close lifts its assets under management to about $800m. Casspi is the founder and sole managing partner, and the new fund's backers include Sequoia Capital, US pension funds, leading American universities and family offices, according to reports in Globes and The Jerusalem Post. Calcalist added that Israeli institutional investors also took part.

A quarter of a billion dollars is not a large fund by the standards of a year in which established firms have raised in the billions. What makes the close notable is the combination: a single decision maker, a portfolio built in less than two years, and limited partners who usually prefer teams, track records measured in decades and elaborate governance. The fact that such investors were willing to write cheques to a one-person firm says something about where institutional money thinks early-stage returns are now being made, and what it is prepared to tolerate to get them.

A concentrated book by design

The fund's plan is unusually explicit. Swish expects to back about 12 companies, with an average investment of roughly $20m each, mainly at seed and Series A. It will usually act as co-lead alongside another fund or as a participant in a round led by someone else. A separate Opportunity Fund lets the firm follow its companies from seed through growth stages. Globes lists the earlier vehicles as a $60m seed fund and a $100m Opportunity Fund.

The sectors are cybersecurity, technology infrastructure, physical AI, defence, AI for industrial applications and government technology. That is a list that plays to Israel's strengths in security and defence engineering, and to the current appetite of American buyers and investors for anything that sits close to AI infrastructure.

Two features of the plan deserve attention. The first is cheque size. An average of $20m per company at seed and Series A would have been a growth-stage ticket not long ago. It reflects how the first institutional rounds of AI and security companies have swollen, and how a small fund that wants meaningful ownership has to show up with real money from the start. The second is the arithmetic of concentration. Twelve positions of about $20m add up to roughly the size of the fund, which leaves the main vehicle with little room for follow-on support. The Opportunity Fund structure is the answer to that, but it also means that the firm's results will depend heavily on a handful of names.

The track record behind the raise

Swish's pitch rests on a portfolio that has marked up fast. Globes and The Jerusalem Post report that the firm has invested in 20 companies, eight of which have become unicorns, with a combined portfolio valuation above $60bn. Calcalist put the figures at 21 companies and nine unicorns. Named holdings include Cognition AI, Applied Compute, EON, Wonderful, Upwind, Long Lake, Glow, Tenzai, PointFive, Irregular and Turnout.

Calcalist reported that Wonderful raised $550m at a $5bn valuation and that Upwind raised $300m at a $3.8bn valuation, both more than doubling their valuations in under a year. It also reported that Cognition is raising capital at a $47bn valuation, three months after raising $1bn at $26bn. Casspi told the paper that "the aggregate value of the companies will reach almost $100 billion."

These are impressive numbers, but they are marks, not distributions. No exits have been announced, and the valuations were set in a market where AI rounds are repriced upwards at a speed that makes year-old prices look quaint. Investors should also note that the vehicles disclosed so far, $60m, $100m and now $250m, add up to $410m. The firm has not explained how it reaches the $800m assets figure, and the gap is large enough that it cannot simply be committed capital.

Why Sequoia and pensions back a solo partner

The presence of Sequoia among the limited partners is the most telling detail. Calcalist reported that Sequoia also backed the previous fund. Large firms have long used small managers as a scouting network, gaining early sight of companies they may later lead at higher prices. A solo investor with access to Israeli security founders and a reputation outside the usual venture circles is exactly the kind of channel that a big platform cannot easily build in-house.

For pensions and endowments, the logic is different. They have spent the past few years watching their large venture commitments tied up in illiquid late-stage positions. A small, focused early-stage fund offers the theoretical upside of owning companies before the price explodes, and a manager whose personal brand opens doors with founders. Casspi's own framing stresses selectivity: "We believe in carefully selecting a small number of exceptional founders each year," he said.

The risks of the model are equally plain. A solo general partner concentrates key person risk to the maximum: if the manager leaves, falls ill or loses interest, there is no partnership to fall back on. Decisions are fast but unchecked by colleagues. And a portfolio of a dozen companies leaves little margin for the inevitable failures. Institutional investors have accepted those risks before, usually after a manager has produced a few distributions. Here they have accepted them on the strength of paper gains.

What the close says about the market

Swish fits a pattern visible in this year's venture fundraising. Much of the attention has gone to the largest platforms, which can write growth cheques into AI companies at enormous valuations. The other beneficiaries are small, sharply defined managers who claim privileged access. A generalist fund of a few hundred million dollars without an obvious edge has a harder story to tell than either.

Israel is a natural home for the second model. Its security and defence ecosystem produces founders with deep technical networks, and global buyers have been paying high prices for its companies. A local manager who can co-lead the first institutional round, and then help the company reach American customers and investors, has a clear role. Calcalist quoted Casspi as saying it was important to him to raise money from institutional investors in Israel, which suggests he also sees value in domestic sponsorship.

Our view

Swish's close is less a story about a celebrity investor than about what limited partners are now willing to buy. They are paying for access and concentration, and they are accepting a single decision maker to get it. That works brilliantly when the portfolio includes companies like Cognition and Wonderful, and badly when marks fall.

Founders should read the $20m average cheque as a sign that early rounds in AI, security and defence are being priced and sized like growth rounds, with the expectations that come with that. Investors in such funds should ask for clarity on how assets under management are counted, how reserves are allocated between the main fund and the Opportunity Fund, and what succession plan exists for a one-person firm. What to watch: whether any of the eight or nine unicorns produces a liquidity event, and whether Swish's next raise comes from the same institutions or from a broader base. Paper gains got this fund closed. Distributions will decide the next one.

Sources

• Globes: Omri Casspi's Swish Ventures closes $250m fund
• The Jerusalem Post: Firm led by first Israeli NBA player Omri Casspi opens new $250 million fund for high-tech startups
• Calcalist Tech: Omri Casspi raises $250 million after big wins with Cognition, Wonderful and Upwind
• Ynetnews: Former NBA star Omri Casspi raises $250 million for Swish Ventures' third fund

Filed under Venture Capital. Written by InvestTimes Editorial under our editorial standards. Commentary and analysis for information only, not investment advice.