Fireside Chat, Interview
The New VC Game | CoreWeave Sell-Off Creates Panic | Revenue Multiples Out of Control
- Large venture funds are expected to require billion-dollar positions as standard, with seed checks of $50 million to $100 million deemed insufficient to generate meaningful returns, necessitating at least 51% ownership to achieve such positions in $5 billion to $10 billion sized funds.
- The investment landscape is projected to concentrate capital into fewer names due to a "power law on steroids" requirement, where top-tier VCs will only pursue companies with "trillion dollar" potential, leaving software companies unable to reach massive market caps "wholly unfundable."
- Early-stage investors may need to "carry" portfolio companies lacking hyper-growth metrics or specific "three-box" criteria, replicating "old days" dynamics where seed funds of $50 million to $100 million struggle to achieve 5x returns due to high entry prices and low ownership percentages.
- The market is expected to see a shift where only 10 to 20 decacorns per decade emerge, with a variance suggesting 90 decacorns is an outlier, as the industry cannot exclusively pursue $25 billion outcomes given there are not enough deals to sustain more than five funds.
- Venture firms are anticipated to require $800 million in fee income to achieve a meaningful market cap, with only the very largest firms likely to pursue an IPO, as the entire industry cannot rely solely on "one-and-done" companies which are viewed as a trap for liquidity and dividend distribution.
- The "inference market" is expected to potentially reach $300 billion to $500 billion, with revenue opportunities an order of magnitude larger than current open weight model providers, driving valuation growth for companies like Modal which may see margins rise into the mid-30s and hit 50%.
- Companies like Town are expected to reach a $1.4 billion valuation by year-end and $5 billion the following year, while Andrew's $90 billion valuation is subject to scrutiny against public comps, and the $1 billion valuation mark is predicted to become the new $100 million entry point for Series B.
- Aggregate market capitalization for SpaceX, Anthropic, and OpenAI is projected to shift from $400 billion three years ago to $5 trillion today, creating $4 trillion in value, with Anthropic's IPO priced at $1.5 trillion and growth rates expected to slow from current unsustainable levels.
- The venture market is expected to heat up as LPs recycle capital from SpaceX returns and anticipate IPOs from Anthropic and OpenAI in Q1 or Q2, though the "AI trade" may show cracks if end-user demand slows, causing volatility for neoclouds like CoreWeave.
- CoreWeave is expected to remain a "B-tier asset" sensitive to system shocks and debt loads, with bond prices implying a "lowest of default" risk, whereas "A-tier" assets like NVIDIA are expected to be more resilient despite trading at a premium.
- Code review is expected to become the single most important software category, as AI-generated code requires an order of magnitude more time to review and make production-ready than to build, creating a massive opportunity for companies like Replit.
- Investors are expected to route traffic based on price and quality, with Replit continuing to use OpenAI models because they offer sufficient quality parity while being 30 to 40 percent better in price, even as Anthropic's market share potentially drops 10% from current levels.
- Berkshire is expected to transform "dull" software assets into enduring holdings similar to See's Candy or Coke for a decade or longer, while Harvard's endowment is anticipated to continue driving returns from venture allocations with a specific "banger 27" IRR.
- Elon's strategy of gifting a share of SpaceX to every American kid is expected to generate an "Elon premium," reduce toxicity around wealth creation, and create a loyal shareholder base, potentially leading to a Tesla-SpaceX merger following predictions that China will dominate EVs and human driving.
- Andreessen is expected to go public, though this may shift the firm's culture toward being more risk-averse, and the industry will see "many more firms IPO" only if they reach a critical mass of fee income, as the "B outcome" of investment memos requires a billion-dollar position.