Carter Steffes

x.com/i/article/2046…

Monthly VC/LP debrief. What I actually saw in March: 1/ Spoke with a multi-billion fund. Their thesis is simple: fund size follows the size of the prize. When @SpaceX is tracking toward $1T still private, a $300M fund can't matter. They raise in months, not years, and build IR teams like corporates. The gap between how mega funds operate and how everyone else does is wider than most people think. (h/t @jkhamehl) 2/ The Midas List doesn't look the way most people expect. Only ~35–42% have operator/founder backgrounds. ~25–35% came from finance. The rest from engineering, journalism, career VC. Great investing is about temperament, not resume. (via @kwharrison13) 3/ @cartainc's data across 2,904 US funds is sobering. Most vintages are at or below water on TVPI. DPI for anything post-2019 is near zero. The 2021 vintage top decile sits at 1.71x TVPI – with virtually no distributions. Even the 2017-18 top 10% returned only 1.35x DPI after 7-8 years. That's the actual bar. (via @PeterJ_Walker) 4/ The LP market is bifurcating fast. Late-stage AI co-invest has unlimited demand, but is full of unsanctioned SPVs and fees that can kill returns (one example seen: 15% upfront + 20% carry + 30% over 2x). Meanwhile, the top 10-20 multi-stage brands are oversubscribed. For emerging managers without a spinout story or a strong operator brand, it's the hardest fundraising environment in a decade. (via @Samirkaji) 5/ LP selection is portfolio construction in reverse. The most expensive LP mistakes aren't the ones who say no – they're the ones who say yes and then behave like owners: demanding co-invest on every deal, quarterly calls, board visibility on 20% of your fund. Rule of thumb: every LP should make Fund II easier to raise, not harder. (my observation) 6/ UTIMCO's disclosures show the power law in its most extreme form: @ThriveCapital's 2022 fund at 126%+ IRR on @OpenAI and @cursor_ai. @notablecap swinging from -48% to 96% IRR in one year on @AnthropicAI. All paper. The caveat nobody says loud enough: this cycle's returns are almost entirely a story of who got into 2-3 AI companies at the right time. (via @maddierenbarger) 7/ One 5x fund doesn't make a great VC. On a $5M fund, a single lucky bet can look like genius. What separates @usv and @foundersfund isn't a blowout vintage – it's consistency across multiple funds, cycles, and GP generations. For LPs, the job is finding franchises worth backing repeatedly, not hunting whoever got lucky last. (via @Lawrence_Ou) 8/ Venture may be having its Nifty Fifty moment. In 1972, Wall Street was right about the companies (Disney, Coke, McDonald's all won) but catastrophically wrong about price. Buyers underperformed the market for 26 years. Now 60%+ of global VC dollars in Q1 2025 went to $500M+ rounds. @OpenAI raised $110B – bigger than any IPO in history. Being right about the company has never been enough. You also have to be right about the price. (via @Jeffreyw5000) 9/ After 25 years of Lean Startup and customer development, startup survival rates haven't improved – and seed-to-Series A conversion has actually collapsed. The culprit is the Red Queen: when everyone runs the same playbook, no one wins with it. Winning requires doing something different, not executing the shared framework better. (via @patrick_oshag) 10/ @RonConway's actual value-add: showing up at Airbnb during COVID, OpenAI during its governance crisis, and across the ecosystem during SVB's collapse – in an hour, when the founder needed it. Not a deck. Not a platform slide. Thirty years of compounding relationships where the people he backed in 2005 are the ones who matter in 2025. That's the seed fund model that actually works. (via @jaltma) Every month I track new fund launches, LP events, market reports, and what's actually moving in VC/LP. All of it in the @murphcapital newsletter: murphcapital.substack.com/p/edition-11-3…