On Wednesday, the New York Times reported that OpenAI is considering delaying its initial public offering until 2027. CEO Sam Altman has told advisers he wants a valuation of $1 trillion, up from the $852 billion post-money valuation in its March fundraise. His advisers presented two options: wait for the market to stabilize and let the company's financials grow into a trillion-dollar number, or accept a lower valuation and list by late 2026. Altman chose to wait.
The filing will be interesting reading. In the first quarter of 2026, OpenAI generated $5.7 billion in revenue and posted a net loss of $21.3 billion. The company projects spending $50 billion on compute this year alone — a figure its president Greg Brockman confirmed during the Musk-Altman trial. It has committed to $600 billion in infrastructure spending by 2030, down from an earlier target of $1.4 trillion but still roughly 24 times its current annualized revenue of more than $25 billion.
The Loop
What is unusual about this delay is its cause. OpenAI is not waiting because its business is weak. It is waiting because the market its spending creates is hostile to its listing. OpenAI's compute budget flows to hyperscalers. The hyperscalers buy memory. DRAM contract prices surged as much as 98 percent in the first quarter of 2026, the steepest quarterly increase on record. Micron reported $41.5 billion in quarterly revenue on Wednesday, more than four times its year-ago figure, and surged 16 percent.
The same memory that makes OpenAI's models possible makes Apple's hardware more expensive. On Wednesday, Apple announced 17 to 25 percent price increases on MacBook and iPad, citing unprecedented component costs. Apple shares fell 6 percent, erasing $265 billion in market value. The Nasdaq posted its fourth consecutive daily decline. South Korea's KOSPI triggered its fifth circuit breaker of the year. The index that OpenAI would list on is falling in part because of the spending OpenAI requires to operate.
The Precedent
OpenAI's advisers are watching SpaceX. The rocket company went public on June 12 with a $75 billion raise at $135 per share — the largest initial public offering in history. It opened at $150 and briefly pushed past both Amazon and Microsoft in market capitalization. Then the tech selloff arrived. SpaceX peaked at $225.64 on June 16 and fell to $147.11 by June 23, briefly trading below its opening-day price and erasing more than $600 billion in market value. In eleven trading days it went from the most successful debut in capital markets history to a cautionary tale. If SpaceX can lose that much value that quickly, a company burning $21 billion per quarter cannot rely on first-week enthusiasm to sustain a trillion-dollar valuation.
The Proxy
The sharpest damage was five thousand miles from San Francisco. SoftBank Group fell as much as 13 percent on Friday morning in Tokyo, its sharpest single-day decline in months. SoftBank has committed roughly $65 billion to build a 13 percent stake in OpenAI. On that thesis, SoftBank overtook Toyota as Japan's most valuable company earlier this month, the first time in 23 years that anyone displaced the automaker from the top of the Japanese corporate hierarchy. The anticipation of OpenAI's public debut had become SoftBank's market value. The delay does not reduce OpenAI's intrinsic worth. It postpones the moment when the private mark becomes a public price. For a company whose capitalization is built on the expected realization of an unrealized gain, postponement is indistinguishable from loss.
The window OpenAI needs is the window OpenAI is closing. Its compute spending drives the memory shortage that raises hardware costs that triggers price increases at consumer technology companies that sells off the Nasdaq that scares its own advisers into recommending delay. The company at the center of the AI boom has become the mechanism by which the boom undermines itself in public markets. Altman can wait for the storm to pass. He cannot stop being the storm.