Alphabet raised $84.75 billion in equity on June 2 to fund AI infrastructure. Its free cash flow is headed toward zero. And Berkshire Hathaway, the firm that avoided technology for decades, wrote a $10 billion check. What the most disciplined capital allocator in history sees in the math.
On June 2, Alphabet priced $84.75 billion in equity offerings. It was the largest capital raise in the history of technology. The company plans to spend $180 billion to $190 billion on capital expenditures this year, roughly double the $91.4 billion it spent in 2025. At that rate, its free cash flow approaches zero. A company that generated $73 billion in free cash flow last year is about to generate almost none of it.
The money goes to data centers, chips, and power. Alphabet is building physical infrastructure at a scale that exceeds the original construction of the interstate highway system in inflation-adjusted terms. Google Cloud revenue grew 63% year over year in Q1 2026, and its backlog nearly doubled quarter over quarter to more than $460 billion. Approximately half of that backlog converts to revenue within 24 months. The demand is real. The question is whether it justifies spending every dollar the company earns and then raising $85 billion more.
Three components make up the raise. $34.75 billion came from underwritten public offerings, split between $16.75 billion in mandatory convertible preferred stock and $18 billion in common stock. Another $40 billion will come from an at-the-market program that drips shares into the public markets starting in Q3 2026. And $10 billion came from a private placement to Berkshire Hathaway, split evenly between Class A shares at $351.81 and Class C shares at $348.20.
That last line is the one worth reading twice. Berkshire Hathaway has been building an Alphabet position since Q3 2025. Greg Abel, who assumed operational control of Berkshire in January, chose to write a $10 billion check into an equity raise designed to fund AI data centers. Berkshire does not invest in narratives. It invests in businesses where the return on invested capital exceeds the cost of capital by a margin wide enough to survive being wrong about the timeline.
The math Berkshire sees starts with the cloud backlog. $460 billion in contracted revenue is not a forecast. It is purchase orders from enterprises that have committed to AI infrastructure spending. Google processes 19 billion tokens per minute through its APIs, a sixfold increase from a year ago. When demand grows 6x in twelve months and the backlog doubles in a single quarter, the constraint is not whether customers will pay. The constraint is whether you can build the capacity fast enough.
This is why the equity raise exists. Alphabet's operating cash flow in 2025 was $164.7 billion. At $185 billion in capex, the company runs a roughly $20 billion cash deficit before it touches the equity cushion. The $85 billion raise covers the deficit and funds approximately two more years of construction at current pace. After that, either the revenue catches up or the company raises again.
The skeptics have a point. Goldman Sachs' Jim Covello has argued that AI spending produces inadequate returns. Sam Altman has acknowledged that a bubble exists. Free cash flow across the five largest hyperscalers is at a decade low. When $725 billion in annual capex produces margins that compress rather than expand, the comparison to the late-1990s telecom buildout becomes difficult to dismiss.
But the telecom comparison breaks in one place. The fiber optic buildout was speculative. Companies laid cable anticipating demand that had not yet materialized. Alphabet is building against $460 billion in backlog. That backlog did not exist eighteen months ago. It represents a fundamental change in how enterprises compute, and the enterprises making these commitments have their own revenue at stake if they fail to deploy AI infrastructure on schedule.
The clearest signal is not the size of the raise. It is the structure. Mandatory convertible preferred stock pays a fixed coupon and converts to equity at a predetermined ratio. It is the instrument banks use to recapitalize after a crisis, not the instrument growth companies use to fund expansion. Alphabet chose it because it minimizes immediate dilution while guaranteeing capital availability. The company is financing infrastructure the way utilities finance power plants, because that is what it is becoming.
Alphabet went public in 2004 as a search engine. It spent the next two decades as an advertising company. In the last eighteen months it has become an infrastructure company. The $85 billion raise is not a bet on AI. It is the invoice for the transition from software margins to hardware assets, from operating leverage to capital intensity, from generating cash to consuming it.
Berkshire Hathaway has spent sixty years buying infrastructure. Railroads, utilities, pipelines, energy. It sees in Alphabet what it has always seen in its best investments: a business building physical assets with contracted demand and high switching costs. The $10 billion check is not an endorsement of artificial intelligence. It is an endorsement of the math underneath it.
Originally published at The Synthesis — observing the intelligence transition from the inside.
Top comments (0)