$852 Billion Later, OpenAI's IPO Math Still Doesn't Add Up

The Numbers Game

$852 billion. That's the post-money valuation OpenAI achieved after closing its record-breaking $122 billion funding round last month. For context, that makes the ChatGPT maker more valuable than Tesla, Meta, or Berkshire Hathaway. It took Apple forty years to reach comparable heights. OpenAI managed it in three.

The scale defies comprehension. Just the funding round itself — $122 billion — exceeds the GDP of Ukraine. Amazon anchored the deal with $50 billion, while Nvidia and SoftBank each contributed $30 billion. The remaining $12 billion came from a constellation of institutional investors, plus $3 billion raised through bank channels from retail investors for the first time.

Revenue vs. Reality

OpenAI's defenders point to explosive growth. The company hit $2 billion in monthly revenue by March, translating to a $24 billion annual run rate. It boasts 900 million weekly ChatGPT users and claims enterprise customers now represent 40% of revenue, up from 30% last year. Its advertising pilot generated $100 million in annual recurring revenue within six weeks.

But here's the uncomfortable truth: OpenAI loses money. Spectacularly.

The company burns through $14.1 billion annually just on inference costs — the computational expense of running AI models for real user queries. Total cash burn is projected to rise from $17 billion in 2026 to $47 billion in 2028. OpenAI doesn't expect profitability until 2030, with cumulative losses exceeding $200 billion through that period.

The Dot-Com Mirror

Every bubble has its own rhythm, but certain patterns repeat. The current AI frenzy shares DNA with the dot-com crash of 2000, though with crucial differences that make the comparison both illuminating and incomplete.

Valuation multiples for major indices were higher during the dot-com peak. The Nasdaq-100's forward P/E ratio reached 60× in March 2000, versus roughly 26× today. Cisco traded at over 200 times earnings at its peak; Nvidia currently trades at less than 50 times.

Yet concentration levels are unprecedented. By late 2025, the five largest companies held 30% of the S&P 500 and 20% of the MSCI World — the greatest concentration in half a century. When 61% of all global venture capital flows to AI startups, as it did in 2025, markets stop pricing individual companies and start betting on an entire technological paradigm.

The IPO Inevitability

OpenAI's path to public markets appears locked in for Q4 2026, targeting a valuation near $1 trillion. The company has begun informal discussions with Wall Street banks and hired finance executives for investor relations. CFO Sarah Friar suggests 2027 as more realistic, but the groundwork suggests urgency.

The IPO isn't optional — it's necessary. At current burn rates, OpenAI requires continuous capital infusion beyond what private markets can sustain. The company faces a structural challenge: building "gigawatt infrastructure every week" while proving to public shareholders that trillion-dollar valuations reflect economic reality rather than speculative fever.

The Anthropic Factor

Competition adds another layer of complexity. Anthropic, OpenAI's primary rival, raised $30 billion at a $380 billion valuation in February and is reportedly considering its own public debut in late 2026. The company's annualised revenue reached $19 billion in early 2026 — growing at 10× annually compared to OpenAI's 3.4× rate.

At current trajectories, Anthropic could surpass OpenAI's revenue by mid-2026. For a company commanding an $852 billion valuation, that's the kind of competitive pressure that keeps investment bankers awake at night.

The Circular Logic Problem

The most troubling aspect of current AI valuations isn't the numbers themselves — it's how they're generated. Amazon invests $50 billion in OpenAI, then provides cloud services that OpenAI pays for with that same capital. Nvidia contributes $30 billion, securing priority access to its own chips. Microsoft maintains its partnership while routing Azure customers to OpenAI's APIs.

This isn't revenue growth; it's an elaborate shell game. Today's "booming" AI revenue reflects internal recycling of investment capital rather than organic customer demand. The economic profit from external customers remains largely hypothetical.

Betting the House on AGI

OpenAI's valuation ultimately depends on achieving Artificial General Intelligence — the holy grail that transforms AI from productivity tool to societal infrastructure. CEO Sam Altman has committed to spending $1.4 trillion over eight years to reach this milestone. Recent reports suggest the company has scaled back to $600 billion through 2030, but that's still an astronomical bet.

The structural risks are real. SoftBank's $30 billion commitment includes a clawback provision if OpenAI fails to complete its conversion to for-profit status by December 31. Elon Musk's ongoing litigation challenges the conversion process. Regulatory approval from Microsoft, the California attorney general, and courts remains pending.

Market Prediction

The most probable outcome isn't a dot-com-style collapse but a sharp bifurcation. Companies with defensible revenue models and improving unit economics will continue attracting premium capital. Those operating on narrative alone will face increasing pressure as investor patience with unprofitable AI positions erodes.

OpenAI sits uncomfortably between these categories. Its technology is genuine, its user base massive, its growth remarkable. But an $852 billion valuation for a company burning $14 billion annually requires assumptions about future profitability that border on religious faith.

By Q4 2026, when OpenAI likely debuts as a public company, markets will demand answers to questions that venture capital has been content to defer: When will unit economics turn positive? How defensible is the moat against open-source alternatives? Can a trillion-dollar corporation justify its valuation through cash flows rather than PowerPoint projections?

The AI revolution is real. The valuations attached to it remain an elaborate exercise in collective wishful thinking.