Fewer Than 20 of 3,000 US Venture Firms Consistently Return 3x Net, and the Middle of the Market Is Where Firms Go to Die, Accolade's Verdiyan Argues

54 minutes ago

4 min read

What's New

Venture has become winner-take-all at the firm level, and the survivors sit at two extremes: brand-name platforms whose past exits pull in the next generation of founders, and small seed funds with a specific right to win and a fund size they refuse to grow. Aram Verdiyan, Partner at Accolade Partners, argues this in a podcast on How I Invest. Accolade's database of more than 3,000 US firms across 2005 to today shows only about 20 that have consecutively delivered 3x net funds, the hurdle needed to justify 10-plus years of illiquidity. The firms that fail are the ones that scale a $75 million seed strategy into a $500 million fund and land in the messy middle, too big to co-found at Berkeley and too small to beat Sequoia.

Why It Matters

The consensus among LPs is that top-quartile persistence is real and manager selection is a numbers game. Verdiyan agrees persistence exists but says it lives almost entirely at the large end, and that seed persistence requires differentiation most generalist firms cannot show. On the other side are the 2,000 to 3,000 seed funds competing for 15 to 20% ownership in the hottest AI categories, and the spinouts raising mid-sized funds with no structural edge. Accolade runs $8 billion in fund-of-funds and held its own fund at $500 million, and it will take 50% or more of a manager it believes in, so the thesis is also a sales pitch to the GPs it wants to anchor.

Big Picture Drivers

Company dispersion drives fund dispersion: Only 10 to 20 companies a year produce multi-billion exits. To 3x the roughly $200 billion raised annually in US venture, the industry needs $1 to $3 trillion of exit value and meaningful ownership in the handful of companies that create it.
Brand is a flywheel: Past outcomes give large firms the reputation to lead rounds for the next cohort of founders. Verdiyan counts Thrive, Founders Fund, and Andreessen as the only firms in two decades to join that top bracket from a standing start, plus a few solo GPs.
Fund math punishes participation: A $50 million fund owning 5% at entry ends near 2.5% after dilution and returns half its fund on a $1 billion exit. Owning 10% or more, as a lead or co-lead, turns the same exit into multiples of the fund and a $10 billion exit into a 10x fund.
Right to win replaces generalism at seed: Technical GPs co-founding with engineers at Stanford, a decades-long franchise in Israel, and sector specialists who research and co-ideate companies in energy or insurance win rounds even against bulge-bracket term sheets.
AI broke the fundraising rules: Last year 10 firms raised more than 50% of venture capital and 18 raised 80%; five companies absorbed more than half of company-level capital over two years. A $30 billion private round closed in a month.
Marks lag reality: Post-SaaS-repricing, a $300 million company growing 25% at breakeven cannot IPO, is no longer a private equity target, and still sits on the books at 15x revenue. Accolade underwrites managers on contextual valuation, not reported marks.

By The Numbers

20 of 3,000: US venture firms with consecutive 3x net funds across vintages since 2005.
12+ years: Average age of a US unicorn, the duration that sets the 3x hurdle.
$30 to $50 billion: Exit value a large firm needs every vintage to produce a 3 to 5x fund.
10 and 18: Firms accounting for more than 50% and 80% of last year's venture fundraising.
18 to 20: Managers per Accolade fund-of-funds, with positions as high as 50% of a manager's fund.
6: SaaS companies trading above 10x forward revenue at the trough of the repricing, versus dozens two years earlier.

Key Trends to Watch

Seed funds refusing to scale: Verdiyan's standard ask to a winning $75 million manager is to raise another $75 million, with Accolade taking all of it. Watch whether the best seed firms hold size or follow demand into the messy middle.
Concentration limits rising: Accolade backed a fund one with a 40% single-deal limit without pushback. Expect more LPAs to permit 20 to 30% of a fund in a runner.
Late stage as a distinct venture category: Series M rounds and 12-year private lives mean venture-like returns now exist at mid and late stage. Watch whether allocators formalize it as a separate bucket.
Experimental ARR in AI diligence: Companies annualizing three-month pilots and raising on the result will show up as plateaued portfolios. Watch renewal cycles, not first-year run rates.

Memorable Quotes

"Brand begets the top deals. The top deals preserves a brand." The flywheel that makes persistence a large-firm phenomenon.
"There is usually a specific reason they're not re-upping with you." Why a 50% anchor LP declining a fund is a symptom, not the cause, of a manager's problem.
"AI has broken the rules of fundraising, valuations, company traction and everything around there." Verdiyan's explanation for the most confusing venture market of his career.
"We're not the smartest LP out there. We have a singular focus." The discipline that kept Accolade's own fund at $500 million.

The Wrap

The thesis holds if the next decade's consistent 3x net funds keep coming from the same two extremes, brand platforms and size-disciplined specialists, and if firms that grow from $75 million to $500 million show the mean reversion Verdiyan predicts. It also holds if lead and co-lead ownership keeps separating fund returners from participatory portfolios outside the AI mega-rounds. It fails if AI creates enough multi-billion exits that mid-sized generalist funds hit 3x through participation alone, or if a wave of new challengers proves the top bracket is more open than three firms in twenty years suggests. The 2021 to 2024 seed vintages, raised at the peak of seed competition and now maturing through the AI repricing, will settle the count.