Technology stocks are crashing as investors confront an uncomfortable reality: the artificial intelligence boom requires staggering capital expenditures that may never justify their costs. Asia-Pacific shares fell 0.16% on October 9, and the region appeared headed for a second straight weekly decline exceeding 1%, as markets grappled with elevated energy prices, surging bond yields, and the huge funding rounds required to buy AI chips.
The immediate catalyst was brutal. OpenAI’s annualized revenue came in at US$20 billion—significantly below what the company had previously signaled to investors. The disappointment rippled across Wall Street’s technology sector, which dragged broader indexes lower overnight. This isn’t simply a valuation miss. It raises a fundamental question: do AI infrastructure economics work at the scale companies are betting billions on?
The Capital Problem Becomes Real
SpaceX, Broadcom, and Oracle are each preparing to raise billions of dollars specifically to purchase high-end AI chips. Australia’s Firmus, a data-centre operator backed by Nvidia, shelved its US$5 billion initial public offering entirely, citing market volatility and conditions as its reason for returning to private fundraising. When companies with substantial backing can’t access public capital at acceptable valuations, markets are sending a message about confidence.
Chris Weston, head of research at Pepperstone, observed that technology, AI infrastructure, and semiconductor stocks have been “a sea of red.” He noted that “for now, though, the price action suggests investors are becoming more selective about where they want exposure and, importantly, what price they are prepared to pay for future growth.” Selectivity is a polite way of saying investors won’t fund unlimited spending on speculative returns anymore.
The bond market’s revolt compounds the problem. A months-long global sell-off has pushed borrowing costs to multi-decade highs. The U.S. benchmark 10-year Treasury yield sits at 5.226%, near the 24-year high it reached on October 7. Charu Chanana, chief investment strategist at Saxo, offered a clear assessment: “With long-term yields back around multi-decade highs, investors no longer have the luxury of valuing AI growth in a low-cost-of-capital world.”
Why Assumptions Are Unraveling
For years, tech companies assumed they could fund expansion at rock-bottom rates. That era has ended. Higher sovereign yields and increased corporate issuance to fund AI infrastructure mean capital is becoming “both more expensive and more selective, which puts balance sheets and the quality of future earnings firmly in focus,” Chanana added. Suddenly, profitability matters again. Cash flow matters again. Balance-sheet quality matters again.
Geopolitical risk adds more uncertainty. Brent crude futures traded at US$103.70 a barrel during Asian hours on October 9, after rising more than 4% in the previous session. Concerns about the Middle East conflict have fueled inflation worries and higher interest rates globally. President Donald Trump stated on October 8 that the United States wouldn’t launch an attack on Iran before the November midterm elections, but traders remained skeptical about whether the war was truly nearing an end. Nick Twidale, chief market strategist at ATFX Global, flagged the core risk: “The big question for markets is whether Trump sticks to his word if Iranian attacks intensify. Any indication that the White House is reconsidering military action could see oil prices spike sharply higher, particularly with tanker traffic through the Strait of Hormuz already under significant pressure.”
Higher energy costs flow directly into data-center operating expenses. Energy prices are a lever nobody can control. Add elevated rates and geopolitical uncertainty, and the unit economics of AI infrastructure investments look far less compelling than they did six months ago.
Currency and Debt Contagion
Europe’s challenges are moving into focus. The dollar strengthened against the euro, which was set for a fifth straight week of losses. The euro last traded at US$1.122, around the 17-month low it touched earlier in the week amid concerns about French debt. Investors closely watched France’s debt load, budget deficit, and political outlook ahead of the 2027 presidential election. ING analysts warned that “with the two rounds of the presidential elections only in April and May next year, French bond spreads are still looking at more than half a year of higher volatility and continued spread elevation.” France took a hard hit from the broader bond sell-off.
U.S. Treasury markets showed relative stability after solid auctions that week, but ING analysts cautioned that “it does not mean the bond market sell-off is over, but it does at least put it on pause, at least just for now.” This isn’t reassurance. It’s a temporary reprieve.
Japan’s Nikkei fell more than 1%, while markets in South Korea and Taiwan closed for a holiday. The broader Asia-Pacific picture reflects a global reality: when capital becomes expensive and scarce, investors scrutinize speculative technologies brutally.
Why This Matters:
The AI boom rested on two assumptions that are now cracking simultaneously: capital would remain cheap and abundant, and revenue growth would eventually justify unlimited infrastructure spending. Neither holds anymore. Rising bond yields mean companies can’t refinance at yesterday’s rates. OpenAI’s revenue miss shows that growth isn’t automatic. Geopolitical risks and energy price volatility also bring costs companies can’t predict or control. The market isn’t rejecting AI itself—it’s rejecting the business case for funding it at any price. As capital becomes both expensive and selective, only companies with genuine profitability and strong balance sheets will access it. The rest face a reckoning. This correction shows what happens when government stimulus and cheap credit unwind: speculative investments collapse, and reality returns.