AI Chip Stocks Lose $1 Trillion as Intel and Broadcom CEOs Warn of a Looming Funding Crisis

Semiconductor stocks have lost more than $1 trillion (£752 billion) in market value as investors question whether the artificial intelligence boom can deliver returns to justify its enormous costs. The Philadelphia Semiconductor Index fell about 12% over the past week, including a 4.3% drop on 14 October.

Arm shares fell nearly 20%, Applied Materials declined 15% and Intel lost 11%. Broadcom traded 28% below its 52-week high. The sell-off reflects elevated stock valuations, persistent inflation and expectations of higher interest rates, alongside concerns about the funding required to expand AI infrastructure.

Capital Could Become AI's Biggest Bottleneck

Speaking at the DBS Asia Future Forward Conference in Singapore, Broadcom CEO Hock Tan and Intel CEO Lip-Bu Tan warned that access to capital and rising borrowing costs could become greater obstacles to AI expansion than physical shortages.

The industry has faced constraints involving wafer fabrication, memory chips and electricity. However, the executives expect these bottlenecks to ease by 2028 or 2029 as manufacturing capacity expands. Financing remains a longer-term challenge.

Building data centres with multiple gigawatts of computing capacity requires sustained investment. If borrowing becomes too expensive or funding dries up, projects could stall even when the necessary hardware is available.

Companies Must Prove AI Spending Pays Off

Hock Tan criticised speculative investment in AI, comparing some investors to casino gamblers who set extreme expectations without fully understanding the assets involved. He also likened corporate AI experiments to a child spending $10 (£7.52) at a video arcade in 10 minutes without learning anything.

Intel CEO Lip-Bu Tan said higher borrowing costs would increase pressure on companies to demonstrate a clear return on investment. Businesses that cannot show tangible benefits may face demands to cut spending.

If AI models stop improving rapidly or adoption fails to generate sufficient revenue, infrastructure projects could be delayed, reduced or cancelled.

Technology companies and major cloud providers issued $225 billion (£169 billion) in corporate bonds during the first half of 2026, a 974% increase year over year.

However, monthly AI-related debt issuance subsequently fell from a June peak of $113 billion (£85 billion) to $23 billion (£17 billion) in September, indicating a more cautious financing environment.

Companies have also turned to private credit and alternative financing arrangements to fund infrastructure without placing all costs directly on their balance sheets. Broadcom announced an initial $35 billion (£26 billion) financing partnership with Apollo and Blackstone to support infrastructure development for customers including Anthropic.

These arrangements can expose hardware suppliers to financial losses. Broadcom has disclosed lease-related backstop exposure of up to $29 billion (£22 billion). Its five largest end customers account for roughly half its revenue. This leaves the company vulnerable if major clients reduce spending or encounter financial difficulties.

A Funding Squeeze Could Reshape AI

Large technology companies such as Alphabet, Amazon, Meta and Microsoft have established businesses generating substantial cash flow. This gives them greater capacity to absorb higher borrowing costs.

Smaller AI developers often depend on external financing while facing enormous computing and training expenses. Advanced semiconductor fabrication plants can cost $20 billion to $30 billion (£15 billion to £23 billion) each, further raising barriers to entry. A prolonged funding squeeze could therefore make it harder for them to compete.