Oct 30, 2025, 10:56 AM ETFireside Chat, Interview

OpenAI Restructuring: Who Wins and Who Loses & Mercor Raises $350M at a $10BN Valuation

OpenAI has finalized a restructuring deal with Microsoft and the Attorneys General of Delaware and California, resolving a complex legal trap to enable capital formation and a future IPO.
The new corporate structure places a $135 billion Charitable Foundation above the company, which is now a Public Benefit Corporation (PBC) capable of public listing, similar to Patagonia.
Microsoft retains a 27% equity stake in the combined entity, having secured a 10x return on its $13 billion investment plus forward contract rights and Azure business preferences.
Employees hold approximately 20%+ equity, and the charitable foundation holds a similar 20%+, while other investors like SoftBank hold the remainder (approx. 10% for SoftBank).
A significant anomaly in the deal is that CEO Sam Altman holds zero equity shares in the combined entity, a rare arrangement contrasting with the valuation demands of competitors like Elon Musk.
The restructuring unlocks potential for a $2 trillion+ IPO, which speakers anticipate could be the most popular retail IPO in history, potentially raising an additional $200 billion in capital if needed.
Secondary market activity shows conflicting valuations, with SoftBank executing a $500 billion buyback alongside a $300 billion direct investment, creating an arbitrage opportunity for early investors.

Andreessen Horowitz (a16z) raised $10 billion, split into $6 billion for growth, $1.5 billion for AI apps, $1.5 billion for AI infrastructure, and $1 billion for defense.
The scale of the a16z funds represents a "Red Army" strategy in venture capital, where "quantity has quality" and massive check sizes allow firms to outbid smaller competitors on follow-on rounds.
Competitors argue that a16z's "wall of news" and media dominance, rather than just capital, tilt the playing field, forcing smaller firms to work twice as hard for similar returns.
The fund structure allows for "optioning" strategies at Series A and B, where large funds can pay higher prices for follow-on rights, amortizing losses across a massive portfolio while securing the biggest winners.
Critics note that while a16z's strategy appears successful, the primary beneficiaries are the firms themselves, with potential for diminishing returns for LPs if the mega-funds underperform relative to their size.
LPs face "bundling" requirements, often needing to commit to multiple funds to gain access to the preferred early-stage or AI-specific funds.

Merkur raised $350 million at a $10 billion valuation, led by Felicis, achieving $500 million in ARR in just 17 months by providing high-end human feedback (RLHF) for AI model training.
The Merkur case highlights a trend of "private as public" companies where late-stage valuations compress multiples (20x ARR) despite massive growth, mirroring public market dynamics.
Ramp is raising at a $30 billion valuation; however, speakers critique this as a "micro-up round" (22% increase from $23B) that offers minimal IRR compared to earlier-stage explosive growth.
Late-stage venture investing is redefined as "private public" assets, where annualized returns are expected to match mid-cap public stock growth (11%) rather than the 3x-10x multiples of early-stage venture.
High growth rates (e.g., Merkur's 5x) allow companies to grow into high multiples, but investors must underwrite returns based on a 3x target (e.g., $30 billion exit for a $10 billion entry), which is increasingly difficult at ultra-late stages.
The "Spray and Pray" strategy is deemed less effective at Series B and later stages due to the high cost of failure; successful firms like Carta data suggest that only 20% of Series Bs return >5x, necessitating rigorous picking.

Synthesia reportedly turned down a $3 billion acquisition offer from Adobe; the hosts argue founders should generally sell unless certain of a $10-$20 billion public outcome.
The hosts discuss the "IPO or bust" dilemma, suggesting that missing a $3B offer in favor of an IPO carries significant risk if the founder is not prepared for the "crushing weight" of public company management.
The FTC's blockage of the iRobot (Roomba) acquisition by Amazon is cited as a negative outcome of antitrust enforcement, potentially forcing a strong consumer hardware company into bankruptcy.
Prolonged antitrust reviews create "duration risk," pushing founders toward secondary sales rather than waiting for M&A closures that may take 18+ months.
Amazon's recent 10% white-collar layoffs and AWS cloud share struggles (falling from 30%+ to 38% market share with AI competition) highlight the operational risks of legacy tech transitions.
Jeff Bezos's departure is framed as excellent market timing, avoiding the "psychic pain" of navigating the difficult AI cloud transition, while Sundar Pichai remains CEO at Google during its AI pivot.

In a direct comparison, the hosts prefer owning Brex at a $13 billion valuation (700M ARR, 50% growth) over Ramp at $30 billion (1B ARR), prioritizing lower entry multiples despite lower absolute revenue.
Andreessen Horowitz is praised for operational excellence and scaling a platform that allows partners to lead specific "fiefdoms" (e.g., crypto, AI, defense), retaining top talent.
The "Wall of News" strategy used by a16z is identified as a critical differentiator that creates an unavoidable competitive moat against smaller funds.
The "optioning" strategy is highlighted as a key advantage for mega-funds, allowing them to treat early-stage investments as call options on massive follow-on capital deployment.
The consensus is that the "new normal" for venture involves massive fund sizes ($10B+) and concentrated "picking" strategies, as simple "spray and pray" no longer yields sufficient returns at later stages.