Ranking Venture Investors

By Ilya A. Strebulaev (Stanford Graduate School of Business) and Blake Jackson (The Ohio State University) · ilyastrebulaev.substack.com, May 26, 2026 · free to read

A preliminary, deliberately transparent attempt to answer the question venture capital has never answered in public: which firms and which individual investors are actually good at this? The authors build a six-factor scoring method, run it over more than 230,000 investments by 13,000-plus US VCs at 5,000-plus firms across roughly 53,000 companies, and publish 2023 rankings of the top 100 firms and the top 100 individuals. Then they compare the result to the Forbes Midas List and find almost no agreement.

Both lists are transcribed in full, rank by rank, in wiki/attachments/ranking-venture-investors/top-100-lists.md, alongside the post’s own figures and the full article text.

The claim that makes the rest matter

“By our estimates, the top 5% of VCs account for roughly 90% of the industry’s profits.”

This is the Power Law stated about people rather than about portfolio companies — and it is why the authors think a credible ranking is not a vanity exercise. If nine-tenths of the industry’s profit sits with one-twentieth of its practitioners, then a founder choosing an investor, an LP choosing a fund, or a young investor choosing where to build a career is making a decision whose payoff is dominated by whether they landed inside that 5%. The authors’ complaint about the existing answer is blunt: Midas “is largely a black box, and the rankings it produces are hard to reconcile with the underlying record.”

They set the stage with a second concentration argument — seven of the ten most valuable US public companies were venture-backed, and those seven hold 85% of the group’s market capitalisation (the exceptions being Broadcom, Berkshire Hathaway and Walmart). Among US public companies founded since 1979, venture-backed firms are half the count and more than nine-tenths of the R&D.

Top 100 US VC firms — 2023

Sequoia ranks first, with roughly 1.5× the points of second-place DST Global. The rest of the top 15, in the authors’ order: Accel, Andreessen Horowitz, Tiger, Founders Fund, Index, ICONIQ, NEA, General Catalyst, Bessemer, General Atlantic, Sutter Hill, Lightspeed, and Ribbit.

Places 16–30 run Benchmark, Redpoint, Thrive, Notable, Kleiner Perkins, Insight Partners, IVP, Greenoaks, OrbiMed, Coatue, Battery, Paradigm, Venrock, Menlo, Greylock. The tail of the list is where the method shows its teeth — seed and sector specialists (Initialized, Forerunner, QED Investors, Lux, DCVC, Third Rock, Flagship, Versant, Foresite) score alongside crossover and growth shops, and growth/PE firms such as General Atlantic are included by design because they “make considerable investments in VC-backed companies.”

Top 100 individual US VCs — 2023

First is Yuri Milner (DST Global as of 2022), then Alfred Lin and Michael Speiser (Sutter Hill). The rest of the top 15: Meyer Malka (Ribbit), Douglas Leone (Sequoia), Hemant Taneja (General Catalyst), Michael Moritz (Sequoia), Jan Hammer (Index), Scott Sandell (NEA), Peter Thiel (Founders Fund), William Ford (General Atlantic), Joshua Kushner (Thrive), Patrick Grady (Sequoia), Benjamin Horowitz (Andreessen Horowitz), and Keith Rabois (Founders Fund).

Three structural findings sit underneath the list:

• A handful of companies anchor everything. Across the top 100 investors there are only 65 distinct “most important” companies; Snowflake and Robinhood each lead for five different VCs. The companies that generate the most points for their investors are Coupang (over 23,000 points), Kuaishou, Pinduoduo, Snowflake, Rivian and DoorDash.
• Firm strength and individual strength are not the same thing. The top 100 individuals work at just 54 firms — seven each at Sequoia and Andreessen Horowitz, six at Accel — yet 49 of the top 100 firms place nobody in the individual top 100.
• The floor is very low. The average individual VC scores about 18 points; the top 100 average roughly 1,000, more than fifty times as much. Almost 75% of companies in the sample had investment costs exceeding the value of their VCs’ stakes.

The six factors

1. Valuation. A private post-money valuation and a public market cap are not the same number. Convertible preferred stock carries features beyond the conversion option, so post-money systematically overstates implied market cap — Gornall and Strebulaev (2020) put the gap at nearly 50% on average for unicorns. Private valuations are marked down uniformly. See Valuation.
2. Dilution. Each subsequent round erodes an earlier stake. Two $1B exits are not equal: a single-round company leaves its 10% investor at 10%, while three further rounds at 20% dilution each leave that stake at 10% × (1 − 20%)³ = 5.12% — half the value for an identical outcome.
3. Net profits. Subtracting cost separates $10M into a $200B outcome from $10B into a $200B outcome: identical gross profit, wildly different net. This penalises deploying large sums to land few winners.
4. Value add. Extra points where the investor held a full board seat or led the round — the roles that imply real involvement.
5. Human capital decay. The factor the authors call the most important for practitioner relevance: a three-year half-life. Value created in 2022 is worth ~$0.50 in 2025, $0.25 if created in 2019, $0.125 if 2016. Two investors who each turned a 20% stake into a $10B IPO look identical until you date the entry — 2005 versus 2020 is an IRR of 28.7% versus 141.9%, on human capital twenty-one years old versus six. Their answer to “but experience compounds” is that if an old win genuinely sharpened someone, the better deals it produced are scored on their own. See Human Capital.
6. Value allocation. Because individuals move between firms, credit splits. When Mary Meeker left Kleiner Perkins to found Bond Capital in 2018, her Kleiner investments stayed with her as an individual, but for firm rankings they split 25% to Kleiner and 75% to Bond.

Eligibility is narrow on purpose: institutional US-based firms with VC activity, so accelerators, angel groups, corporate VC arms, non-US firms and family offices are out; individuals must have been full-time investment professionals at a qualifying firm in the year before the ranking year.

What the data is built on

The distinctive ingredient is not the commercial data (PitchBook, VentureSource, Preqin, Jay Ritter’s IPO dataset) but a hand-reconciled archive of historical VC firm websites — more than three million URL requests spanning 1996 to 2026, cleaned by hand for some 1,600 firms. It adds roughly 17% more investment affiliations than commercial sources alone and covers the firms responsible for 96% of US VC capital. That is how you recover who was actually at a firm when a deal was done, which is the whole game for ranking people rather than funds.

The candid limit: for the average company in the sample, about 25% of a firm’s points are attributed to “anonymous” individual VCs the authors cannot yet identify. They are soliciting verifiable submissions.

The Midas comparison

This is the sharpest section. Against the US-based VCs in the 2023 Midas List:

• Rank correlation among investors on both lists is about 0.27.
• Midas misses 58 of the authors’ top 100 outright, including one of their top 10 (Yuri Milner).
• Named absences from Midas with obvious records: William Gurley (53rd here; early Uber, first invested 2011, IPO 2019), Gregory McAdoo (39th; early Airbnb in 2009 at Sequoia, IPO 2020), Joshua Kushner, Marc Andreessen, John Walecka.
• They then tried to reverse-engineer Midas from its disclosed methodology. Even the best-fitting replication leaves 49 of its own top 100 absent from the actual Midas list, including four of the top 10.
• Milner, Andreessen and John Doerr have each been affiliated with more IPOs than almost all 73 US-based VCs Midas does include.

Their conjectures for the gap: Forbes may over-weight data from firms that choose to submit it (selection bias), may apply undisclosed eligibility criteria, may make more than sporadic manual adjustments, or may use a consistent component it simply does not disclose.

Caveats worth keeping attached to the numbers

The post says twice, in the authors’ own words, that these are preliminary results published to solicit feedback ahead of the real release, based on investments and valuations from 2022 or earlier, and that they “must not be quoted as final rankings of either VC firms or individual VC investors.” 2023 was chosen deliberately — recent enough to demonstrate the method on good data, stale enough not to represent current standings. The release schedule they gave: white paper now, 2026 firm rankings in the first half of June, 2026 individual rankings after June 15, then annual releases from 2026 (for 2025), a 25-year time series, and global rankings after that.

The white paper is at docsend.com/view/s/qmakz7ad3v9aqyn9; the academic version is Ranking Venture Investors on SSRN.

Connections

• Power Law — the article’s core claim is the power law applied to investors rather than to companies: 5% of VCs, 90% of profits.
• Venture Returns — where the concentration-of-returns literature lives in this wiki.
• How Do Venture Capitalists Make Decisions — the Gompers/Gornall/Kaplan/Strebulaev survey of 885 VCs; same lead academic, the behavioural counterpart to this ranking’s scorekeeping.
• Valuation and Human Capital — the two factors that do the most work in the method.
• Venture Capital — the hub.
• The top 5% of VCs… (tweet) and 5% of VCs generate 90%… (tweet) — the two saved posts that brought this study into the wiki.

Notes

• Firm-name forms are the post’s own short forms; this wiki’s canonical pages are longer in several cases (Index → Index Ventures, Tiger → Tiger Global Management, Bessemer → Bessemer Venture Partners, Ribbit → Ribbit Capital, Andreessen Horowitz → a16z, Lightspeed → Lightspeed Venture Partners, Greylock → Greylock Partners).
• Person pages have deliberately not been minted for the 100 ranked individuals. Where a page already existed it is linked; the rest are named as plain text.