Someone Made A Big Mistake About OpenAI's $1.5 Trillion Value

OpenAI is reportedly discussing a new funding round at either a $1.2 trillion or $1.5 trillion valuation, a $300 billion discrepancy that underscores uncertainty in private AI-company pricing. The company’s reported annualized revenue run rate is $40 billion and it is not expected to reach profitability until 2030, while rival Anthropic reportedly has a $65 billion run rate and may pursue funding or an eventual IPO at a similarly elevated valuation.

Analysis

The valuation discrepancy is not investable for NYT or PG: neither has a meaningful economic linkage, and any short-term attention effect on NYT is immaterial versus advertising, subscriptions, and broader media multiples. The more relevant signal is that private AI financing remains price-insensitive despite unresolved profitability and cash-burn disclosure. That sustains the funding runway for model builders, but it also delays the moment when customers can demand lower inference prices—a constructive near-term setup for compute and networking suppliers rather than for application software vendors.

Over the next 1-3 months, public-market beneficiaries are the infrastructure providers with contracted or quasi-contracted demand: NVDA, AVGO, ANET, VRT, and hyperscalers selling capacity. A large equity round would reduce counterparty/financing concerns around AI capacity commitments, particularly for MSFT and ORCL, but it does not validate returns on their data-center capex; investors should distinguish backlog growth from free-cash-flow conversion. The key second-order risk is that well-capitalized frontier-model firms compete inference pricing down, pressuring AI software vendors whose valuation assumes durable gross-margin expansion.

Consensus may be treating headline private valuations as confirmation of broad AI equity upside. A more useful read is that capital is concentrating in a small number of model developers, increasing their bargaining power versus software distributors while raising the risk of eventual compute overbuild. The structural catalyst for a reversal is not a lower private-mark mark; it is evidence that inference utilization or enterprise seat monetization fails to keep pace with hyperscaler capex through the next two earnings cycles.

Key Decisions for Investors

• No action in NYT or PG on this item; treat any price move tied to the story as noise rather than a fundamental catalyst.
• Maintain a 1-3 month infrastructure bias: long NVDA or AVGO versus a basket of high-multiple AI application software (IGV proxy) only after confirming next-quarter hyperscaler capex guidance. Thesis is invalidated if MSFT, AMZN, GOOGL, and META collectively guide data-center capex materially below consensus or NVDA supply commentary signals order deferrals.
• Prefer a relative-value long GOOGL/AMZN versus MSFT for 3-6 months if frontier-model competition continues to broaden: their exposure to alternative model ecosystems provides optionality, while MSFT bears greater perception risk from concentrated OpenAI-related capacity. Exit if Azure growth reaccelerates materially relative to AWS/GCP or if disclosed OpenAI-related commitments are demonstrably cash-flow accretive.
• Set an earnings watch on VRT and ANET: initiate only if backlog/order growth remains strong while gross-margin guidance holds. A funding headline alone is insufficient; margin dilution, delivery delays, or receivables growth would indicate AI demand is being financed ahead of end-customer monetization.