Fund of funds — one of those structures that seems obvious until you start doing the math. The idea is simple: instead of choosing one venture or private equity fund, an LP buys a stake in a fund that itself invests in dozens of other funds. The result — instant diversification, access to managers who don't let you in alone, and two tiers of fees you'll have to work through.
The idea: a fund that invests in funds
The model is older than it looks, and its origin is a cautionary tale. The first vehicle to carry the name was launched in 1962 by Bernard Cornfeld's Investors Overseas Services: an offshore fund run out of Geneva that bought units in American mutual funds and sold them worldwide on the pitch "Do you sincerely want to be rich?". It gathered hundreds of millions, then collapsed in 1970 as Robert Vesco looted what was left. The respectable, institutional version came later — diversified pools run by banks and specialist allocators for investors who wanted a whole category without choosing the managers inside it.
Mechanically it is a limited partnership. A general partner — the fund-of-funds team — raises capital from limited partners and commits it to a portfolio of underlying funds, drawing it down over several years as those funds call it. The structure sits next to a feeder fund: a feeder pools investors into one target fund, while a fund of funds spreads them across many.
Varieties of fund of funds
The label covers several different businesses. A venture fund of funds backs early-stage managers and depends entirely on access to the handful of funds that drive the power law. A buyout fund of funds diversifies across larger, lower-variance managers, where the fee drag bites hardest because the dispersion it pays to capture is narrower. A hedge fund of funds — the classic 1990s product — spreads across strategies and was the format hit most visibly by fee compression after 2008. A secondary fund of funds buys existing LP stakes at a discount and shortens the J-curve. The retail version, a mutual fund holding other mutual funds, is mostly a convenience wrapper for target-date and allocation products.
Why LPs need this
Diversification.One venture fund holds 20–30 startups; a fund of funds holds 15–25 such funds — that's ~400–700 companies in the portfolio through one ticket. For venture this is critical: power law distributes returns so that without sufficient coverage the median LP gets 1x or worse.
Access. The best funds — Sequoia, Benchmark, Lightspeed — take new LPs mostly in theory; allocation flows to investors already in earlier vintages. A fund of funds that has held those relationships across cycles can secure a ticket a newcomer cannot. The same demand now feeds a wave of access platforms — Moonfare and the broader investor infrastructure — that resell fractions of brand-name funds to smaller LPs.
Expertise.Manager selection in venture is not the same as in public markets. FoF teams spend years building track records, making LP reference calls, calculating DPI and TVPI by vintages. An LP without an internal team is buying exactly this due diligence.
Double fees and net-net
The underlying funds charge the usual 2% management fee and 20% carried interest. The fund of funds adds a second layer — commonly 0.5–1% in management fees (Callan put the 2024 median near 0.76%) and another 5–10% carry on net gains. The carry is what stings: the FoF takes its cut after the underlying GPs have taken theirs, so the same dollar of profit is charged carry twice. Independent estimates put the total drag at roughly two to four points of net IRR against committing to the same funds directly — the hurdle the carried-interest math has to clear before the wrapper earns its place.
Returns after all fees at both levels are called net-net. They need to be compared not to the gross IRR of the venture industry, but to the net IRR of a specific vintage LP — otherwise the numbers don't match up. A typical FoF loses 200–400 bps on the double structure versus direct investing — and that's the price for diversification and access.
Where FoF makes sense
Families without an allocation team. A family office that does not employ three to five senior allocators — people who attend LP conferences, read PPMs, and run reference calls — has no practical way to build a diversified private-markets book alone. A fund of funds rents that capability. For families that build a dedicated vehicle around it, the Hong Kong FIHV and similar regimes set the tax frame.
Venture as tail-allocation.If venture is 5–10% of total allocation and you need to remove single-manager risk, a fund of funds solves the problem with one ticket. For this role double fees are acceptable — the alternative, committing to one venture fund, carries more idiosyncratic risk.
Emerging markets.Local managers in Southeast Asia, Latin America, Africa are often accessible only through regional FoFs — they have relationships with local fundraising and understanding of local regulatory specifics.
Where FoF works poorly
Portfolio overlap.If 15 venture funds hold the same 30 startups (Stripe, OpenAI, Anduril) — real diversification is illusory. The LP pays for access they don't already have through any other fund in the same category.
Blind pool.Unlike direct manager selection, the LP doesn't know in advance which specific underlying funds their money will go to — the commitment is to the strategy and track record of the FoF GP, not to a specific portfolio. For some this is acceptable, for others — a deal-breaker.
Power law dilution.Venture math works so that ~10% of funds collect 90% of returns. FoF holds everything: winners, average, and underperformers. Mean regression suppresses top-decile returns — and the "average" manager through FoF investing turns venture returns into something closer to private equity returns.
Where the model is heading
The traditional fund of funds is under pressure. Private-equity fundraising fell about 13% in 2025 to roughly $480 billion, a third straight annual decline, and global venture fundraising slipped below $100 billion for the first time since 2015. When capital is scarce, LPs concentrate it in the largest established GPs and cut the long tail of emerging managers. That flight to quality runs both ways for a fund of funds: the brand-name funds it sells access to get harder to enter, while the emerging-manager alpha it once harvested gets thinner.
The growth is in the formats that grew out of the model rather than in the model itself. Secondaries and continuation vehicles have become a primary source of liquidity — VC secondaries still trade only a sliver of unicorn value, so the runway is long — and evergreen, multi-strategy products now assemble diversified private-markets portfolios, frequently using secondaries to skip the J-curve. In substance these are funds of funds, sold without the label and, increasingly, with one fee layer rather than two.
If not a fund of funds
An LP with conviction and infrastructure has other routes to the same diversification. Direct manager selection keeps all of the upside but demands the team most families lack. Single-deal SPVs and syndicates — the model behind AngelList and Sydecar — let an investor assemble a bespoke portfolio one position at a time. Co-investment rights granted alongside a primary commitment add exposure at little or no extra fee, and secondaries buy seasoned, partly-drawn portfolios at a discount. Most large allocators now blend several of these rather than hand one GP a second layer of carry.