How VC Funds Make Money: 2 and 20, Carry and the Waterfall

Venture capital funds make money in two ways. The management company charges limited partners an annual fee, usually 2.0–2.5% of committed capital, which pays salaries and running costs. The general partner then takes carried interest, usually 20% of the fund's profits, once investors have their contributed capital back. The fee keeps the lights on; the carry is where partners become wealthy, and it arrives only if the fund has outliers.

The difference between the two streams is easy to miss. A $1 billion fund at 2% earns about $20 million a year in fees whether or not it ever returns a dollar. A $40 million seed fund at 2.5% earns about $1 million a year, enough for two partners at modest salaries. That gap explains a lot about how venture firms behave, including why they keep raising larger funds.

What is "2 and 20" in venture capital?

"2 and 20" is shorthand for a 2% annual management fee and 20% carried interest. The terms are written into the limited partnership agreement (LPA) of a fund, which is usually a Delaware limited partnership. Limited partners (LPs) commit capital. A general partner entity makes the investment decisions and receives the carry. A separate management company employs the team and collects the fee.

Actual US practice in 2024–26 varies around that headline:

Some firms tier their carry, for example 20% rising to 25% above a 3x net multiple. The point is the same: the manager earns more when LPs earn more.

How does the management fee work?

The fee is charged on committed capital, not on capital invested, during the investment period, which typically lasts three to five years. After that it steps down. Over a ten-year life, about 15–25% of commitments goes to fees and expenses unless the fund recycles.

That matters for returns. If $100 million is committed and $20 million goes to fees, only $80 million reaches companies. The portfolio must earn a higher multiple on invested capital than the LP sees on committed capital. In the book's stylized $200 million fund, fees absorb $40 million over ten years, a 3.37x multiple on invested capital becomes 2.69x on committed capital before carry, and 2.35x after.

The fee is also a business. The research companion sketches a $300 million early-stage fund at 2%: about $6 million a year, of which roughly two-thirds goes to compensation, leaving a modest margin. Firms become comfortable only when overlapping funds, raised every two to four years, pay fees at the same time. This is the root of fund-size creep. Fees grow in a straight line with assets, but the number of great early-stage companies does not. A seed manager who triples the fund usually has to write bigger checks at later stages to deploy it. Chapter 3 of The Outlier Equation treats the management company as a business in its own right.

How does carried interest work?

Carried interest is the GP's share of profits. Under a whole-fund (often called "European") waterfall, the order is:

1. Return all contributed capital to LPs, including the capital used for fees.
2. Pay a preferred return, if the fund has one.
3. Give the GP a catch-up, if the LPA provides one, until it has its full share.
4. Split everything after that, usually 80% to LPs and 20% to the GP.

Under a deal-by-deal ("American") waterfall, the GP takes carry as each investment is sold, after returning that deal's cost. That pays partners earlier but creates clawback risk. The research companion's example shows how. A $100 million fund sells Company A in year 3 for $50 million on a $10 million cost, and the GP takes $8 million of carry. The rest of the portfolio disappoints, and the whole fund ends with only $105 million on $100 million contributed. The GP was entitled to just $1 million, so it owes a $7 million clawback. Most LPAs cap the clawback at after-tax carry. If the GP paid 40% in tax, it returns at most $4.8 million, and LPs end up $2.2 million short. That is why most institutional venture funds have moved to whole-fund waterfalls, and why LPs negotiate escrows and guarantees where they have not.

Why do venture funds rarely have a hurdle rate?

Buyout funds usually promise LPs an 8% preferred return before the GP earns carry. Many venture funds do not, and the reason is the shape of venture returns. The research companion's model shows that the hurdle only bites in mediocre funds:

At 1.5x gross the hurdle wipes out the GP's carry, because LPs never earn 8% a year. At 2.0x and above, the catch-up restores full carry, and the hurdle only delays it: in the 3.0x case, carry starts in year 9 instead of year 8. The flip side is the risk LPs worry about. With no hurdle, a GP can earn carry on a return below what public markets paid. Experienced LPs therefore negotiate the terms that bite in mediocre outcomes: hurdles, fee step-downs, fee offsets and clawback security.

How much do fees and carry cost LPs?

The gap between gross and net is wide. The companion's fully reconciled $100 million fund works like this:

The result is 2.45x net TVPI and about 19% net IRR, against about 25% gross IRR. Without recycling, the same manager invests only $81 million and delivers 2.14x net. That 0.3x gap is larger than the difference between 20% and 30% carry at this outcome (2.45x versus 2.27x).

Three rules of thumb follow. A fund needs about 3.0x gross to deliver about 2.5x net, and about 3.7x gross to deliver 3.0x net. A fund that merely returns its cost, 1.0x gross, returns only 0.81x net to LPs, because fees are a real loss. And net TVPI typically runs at about 0.7–0.75 times gross for funds between 2x and 4x. This is why managers pitch portfolio constructions targeting 4–5x gross.

Why do top firms charge 30% carry?

Premium carry is a bet on outliers. At 5.0x gross, 30% carry costs LPs about 0.38x of net multiple compared with 20% (3.67x versus 4.05x). Yet the 30% fund still beats a 20% fund that only reaches 4.0x gross. LPs pay 30% only to managers they believe will produce the outliers, and they accept it because access to those managers is the scarce good.

Benchmark is the best-known example. Founded in 1995 as an equal partnership with small funds, it is widely reported to charge 30% carry. Its record includes a reported $6.7 million investment in eBay in 1997 that became worth billions and an Uber Series A of about $11–12 million at a roughly $60 million post-money in 2011. The premium is justified by a power law: in venture, the difference between a top-quartile and a median fund dwarfs any fee saving. Our explainer on the power law in venture capital shows why.

How does carry reach the people who earn it?

Inside a firm, carry is allocated per fund as "points" of the carry pool. Points vest over about four to six years and reset with each new fund. Their value swings enormously. In the companion's worked example, a $300 million fund's carry pool is zero at 1.0x gross, about $40 million at 2.0x, about $90 million at 3.0x and about $190 million at 5.0x. A point is worth nothing in a mediocre fund and a great deal in an excellent one, which is why senior investors value a carry offer like an option.

Partners must also fund the GP commitment, commonly 1–2% of the fund, with real money. And the tax treatment has conditions. Since the 2017 tax law, Section 1061 requires a three-year holding period for carried-interest gains to count as long-term capital gains. Venture holdings are usually far longer, so the rule mainly affects quick secondary sales and early acquisitions. The legal, tax and fund structuring study covers the details, and the careers and firm organization study covers how carry is split.

How does a VC firm make money over time?

The answer is not one fund but a sequence of them. Returns arrive late: a venture fund typically takes eight to twelve years to return paid-in capital, and the result is usually set by one to three companies. By the time a GP raises its next fund, every two to four years, the previous one has almost no cash returned. LPs therefore underwrite managers on access, judgment and early marks.

That timing produced the 2022–25 squeeze. Carta reported a median DPI of about 0.27x across roughly 2,500 US venture funds in early 2025. US venture fundraising fell to $66.1 billion across 537 funds in 2025, by the PitchBook-NVCA count, while Andreessen Horowitz's reported $15 billion close in January 2026 was more than 18% of all US venture commitments since January 2025, per PitchBook. Fee income concentrated in the largest platforms; emerging managers, living on thin fees and distant carry, took most of the contraction.

Large firms have also built vehicles that change the economics. Andreessen Horowitz registered as an investment adviser in 2019 so it could hold crypto, secondaries and public stock at scale. Sequoia restructured into the evergreen Sequoia Capital Fund in 2021 to hold public winners longer. Our case study on the Sequoia evergreen fund looks at what that changes for LPs. For the full picture of how funds fit into the industry, see our guide to how venture capital works.

Where the money really comes from

In the book's four-factor framing, a venture return is created, priced, governed and turned into cash, and the factors multiply. Fund economics sit across the last two. The fee structure governs incentives: it rewards asset growth whether or not returns follow. The carry structure rewards realization: no carry is paid until LPs have their capital back in cash. A manager whose fund shows high paper marks but no distributions has, in the terms that matter to the GP's own wealth, earned nothing yet. That alignment is the system's main defense against the fee trap, and it works only when fund size stays consistent with the outliers a manager can realistically own.

The takeaway

Venture funds make money from a fee that pays the firm and carry that pays the partners, and the two pull in different directions. Fees reward size; carry rewards outliers that become cash. Fees, expenses and carry turn a 3.0x gross portfolio into about 2.45x for LPs, so recycling, fee step-downs and waterfall design matter more than they look. The best test of any fund is whether its size, fees and carry still make sense if only one company in thirty becomes a true outlier. Chapter 3 of The Outlier Equation works through fees, carry and the LP relationship in full, and Chapter 24 shows how carry shapes careers inside the firm.