Broadcom Plummets Nearly 6% on AI Chip Financing Risk Concerns, Market Cap Falls Below $1.9 Trillion

Broadcom (AVGO-US) plunged 5.94% on Friday (14th), closing at $392.99 per share, with an intraday low of $388.50. Bank of America estimates that financing platforms providing capital to Broadcom's AI chip customers could accumulate $370 billion in debt by mid-2029. Although this debt is not directly borne by Broadcom, the company has guaranteed lease payments for some customers' equipment, sparking market concerns over potential losses.

Broadcom opened at $411.96 and reached an intraday high of $412.50 before selling pressure gradually increased, closing near its daily low. According to Google Finance, the company's market capitalization is approximately $1.87 trillion, with a price-to-earnings ratio of about 65.42.

Reuters reported that Bank of America analyst Tom Curcuruto estimates that if the computing capacity of data centers supported by Broadcom-related financing platforms reaches 20 gigawatts (GW), the platforms' senior debt could rise to $370 billion by mid-2029, with around $150 billion added in 2027 alone.

This model involves financing platforms first raising funds from large institutional investors, then purchasing server racks equipped with Broadcom's custom AI chips and networking equipment, and finally leasing them to AI companies. This allows customers to avoid paying the full setup cost upfront and instead access the required computing power through long-term leasing.

Since borrowing is primarily recorded under the financing platforms' names, the $370 billion debt is not Broadcom's own liability.

However, Broadcom is not entirely detached from this financing arrangement.

To help facilitate funding for these transactions, Broadcom has agreed to guarantee lease payments for some customers' equipment. If AI companies leasing the equipment fail to make timely payments, Broadcom may be required to cover part of the amounts under the agreement.

This arrangement has prompted investors to re-examine the financial foundation behind AI demand. If some customers rely on chip suppliers' guarantees to secure funding for data center construction, it suggests that part of the AI equipment demand is highly dependent on financing rather than being fully supported by customers' own cash flows.

If AI computing demand falls short of expectations, equipment lease rates decline, or leasing customers face financial difficulties, the financing platforms' repayment capacity could be affected, and the guarantees provided by Broadcom could turn into actual losses.