On Monday, Meta Platforms priced two concurrent offerings totaling about $35 billion — $22.5 billion of Class A common stock and $12.5 billion of convertible preferred — to finance its AI infrastructure buildout, with a follow-on sale of up to $60 billion hinted for 2027. The share class was met with a shrug: the stock was up about 1.6% in pre-market. But the detail that deserves more attention is who is expected to absorb the dilution.
The circularity is not new, but it has become conspicuous. Reports that Nvidia, AMD and Broadcom are among the firms interested in buying the new equity turn a routine corporate-finance maneuver into a closed loop: Meta sells shares to fund chip purchases, and the chipmakers that supply the capex buy the shares in return. Each quarter, more of the AI buildout is financed by the vendors themselves, which softens the price signal a stock sale is supposed to send. The market's reaction — a rise, not a decline — suggests investors now read the dilution as a vote of confidence rather than a warning.
For holders of the broader AI trade, the stakes are second-order. If the chipmakers' balance sheets become conduits for hyperscaler capex, the equity market starts valuing AI infrastructure through equity rather than credit — which may be more resilient, but also more interconnected. A setback for any single data center project now shows up on the balance sheets of both buyer and seller at the same time. Meta's $35 billion raise is the cleanest illustration yet of how fast the loop closed.
Image source: i.ibb.co