Naimi: The Capital Crunch Hiding Behind Silicon Valley's AI Boom
Key points
Key takeaways from an Invest Like the Best interview with Ramtin Naimi, founder of Abstract Ventures (November 2026):
The goalposts moved again. Terminal outcomes have shifted from $10B best cases in 2016 to trillion-dollar outcomes today: SpaceX, Anthropic and OpenAI are already three, and Naimi's bet is on two or three more, "not 40 more".
The 'cuck tranche'. Rounds happen faster than ever and losing investors pay 3–5x the primary price to join the same financing via secondary tranches alongside the primary — a necessity for capital-hungry AI companies, not a sign of an unhealthy market.
Markups are the bare minimum. Getting markups no longer feels good, deals are being pulled or renegotiated at a pace not seen since late 2021, and public-market investors are aggressively moving into venture again.
2021 vs now. The pace and FOMO echo 2021, but today's companies are genuinely more interesting — product-market fit is found faster, and founders and their families actually use the products.
Revenue quality caveat. 10–20x year-over-year growth often masks low or negative margins, customer concentration, annualised data revenue, or consumer and usage-based rather than contractual enterprise revenue; margin profiles are nonetheless improving fast — some portfolio companies swing from -15% to +35%.
Labor replacement, not SaaS. Vertical AI companies at 100x revenue multiples look absurd against 13x public SaaS multiples, but the addressable market is labour budgets, not software budgets — law firms spend more on paralegals than on software.
The LP crunch. Mega-funds raised two to three times in 18–24 months, with deployment periods roughly 3x faster (a $2.5B fund deploying since July coming back in January; $3B deployed in nine months and back for $5B); LPs meanwhile get little liquidity outside a handful of names — SpaceX, Anthropic, OpenAI and Databricks — that must generate it for the asset class.
Read more: Invest Like the Best
