a16z: Top 1% Takes Nearly 60% of Venture Capital Profits
Andreessen Horowitz shared a post by partner Jen Kha stating that venture capital exposure and returns are almost not the same game: approximately 57% of the industry's net profits are concentrated in the top 1% of investors, while the top 5% take about 90%, with the remaining 95% splitting the leftover portion. The full article titled "Catching the Venture Capital Train" discusses how the power law permeates the fund layer.
This ratio aligns with calculations from Stanford's Ilya Strebulaev team: about 12,151 investors, over 99,000 investments, with a net profit of approximately $1.2 trillion in 2024 dollars; the top 1% of about 120 people contributes approximately $567 billion, or 56.7%, while the top 5% of about 600 people contributes approximately 90.2%, or $1.08 trillion, leaving the remaining 95% with about $120 billion. In terms of "successful exit counts," the top 1% accounts for about 16.7%, the top 5% for about 45.6%, and about 60% of institutions that have invested at least once contribute exactly zero. The dollar value is steeper than the count.
Kha's article adds to the differentiation at the fund level. According to Accolade data, only about 20 out of approximately 3,000 venture capital firms in the U.S. have consistently achieved 3x net returns over the past twenty years. PitchBook screened 2,143 global funds that reported DPI from 2000 to 2018: 17.0% returned at least 2x, 6.7% returned 3x, and 2.4% returned 5x. In the past decade, the top decile DPI is about 2 to 3 times that of the top quartile and 4 to 10 times the median; in recent years, the top quartile's five-year IRR is about twice the median, with the gap between the decile and quartile at a high observational level. Eight out of the top ten companies by market capitalization in the S&P 500 are tech companies. Other statistics indicate that the top 1% of companies account for about 80% of exit value in 2023 (about 45% if SpaceX is excluded).
The article concludes that diversification fails here; only a few funds have sustained exposure to a few companies. a16z itself manages assets reported to be around $90 billion and is increasing check sizes at the seed stage, effectively using industry statistics to comment on the concentrated strategy of platform funds.
The market mechanism is the allocation of LP quotas. The buy side consists of pension and endowment funds trying to get into that list of 20 firms that consistently achieve 3x; the sell side consists of platforms already on the right side of the power law. Funds flow from 95% of funds to the 1% and 5% of people and the companies they can repeatedly enter. The beneficiaries are institutions that simultaneously hold brand, quota, and follow-on rights; the pressured side consists of long-tail funds with exposure but no top exits. The event-driven aspect is a research summary, not a new fundraise completion.
Buying into the venture capital industry and buying into venture capital profits are separated by a very steep list.
Source: Public Information
ABAB AI Insight
a16z has turned others' labor statistics into its own recruitment brochure. The top 1% taking 57% of net profits indicates that the industry average has no operational significance for LPs; the position to get on board is what matters. Strebulaev treats people rather than funds as the unit, with about 120 names corresponding to over half the profits, showing that partner rights are stronger than brand rights. The fact that 60% of institutions have zero successful exits proves that "participating in venture capital" is one profession, while "participating in profits" is another. The DPI differentiation indicates that paper IRR is no longer sufficient; cash returns are also concentrated.
The capital path is quota compounding. Top funds exchange brand for quotas, use quotas for the next round's allocation, and trade those quotas for the few dozen companies that can contribute trillions in net profits. Out of 3,000 firms, only 20 have consistently achieved 3x, which is the real choice set for LPs. Artificial intelligence further piles exit value onto fewer companies, increasing the cost of missing the train. Platform funds are expanding seed coverage to reduce the probability of missing that 1% of companies, not to allow 95% of peers to share in the profits.
Analogies can be drawn to movie box office, bestsellers, and S&P weighted stocks. The uniqueness of venture capital is that failed projects are counted as zero, while top projects are counted as hundreds of times, making the average mathematically prohibited. The industry phase controls access. Those who can repeatedly appear on that list of 51 funds with 5x returns and that list of 120 people define industry returns; the rest define industry employment.
The structural change is capital concentration. The mechanism is that the return function is extremely sensitive to extreme values, and average holdings cannot replicate those extremes. Diversification is effective in a normal world, but in a power law world, it pins people to the denominator. Venture capital exposure is a ticket, while venture capital returns are a train ticket; the two tickets are rarely sold together.
ABAB News · Law of Cognition
Buying into the industry and buying into profits are separated by a very steep list.
Success in counts does not translate to success in dollars.
Talking about diversification in a power law context is like voluntarily staying in the denominator.