Global Market: AI trade faces biggest test yet as chip stocks tumble across Asia

A selloff in artificial intelligence-related stocks across Asia on Tuesday signalled growing investor concerns about the sustainability of the massive spending boom underpinning the technology sector, as rising bond yields and higher financing costs threaten to undermine the investment case for AI infrastructure, according to a report by Reuters.

Nearly four years after OpenAI launched ChatGPT, billions of dollars have flowed into semiconductor manufacturers, data centre operators, cable companies and other businesses benefiting from an unprecedented race to build computing infrastructure. However, signs of strain are emerging as investors increasingly question whether the expected returns from AI investments will justify the capital being deployed.

Also Read | Global Market: South Korea’s Kospi plunges over 7% as chip stocks tumble on CXMT concerns


According to Reuters, concerns that began surfacing in June have spread across financial markets, with a planned data centre initial public offering in Australia collapsing last week and credit investors becoming more selective about AI-related debt.

In Asia, South Korean chipmakers extended their declines for a fifth consecutive session on Tuesday, accelerating a broader technology selloff. The KOSPI index fell more than 6%, while Japan's memory chipmaker Kioxia plunged 13%, leading losses among AI-linked stocks, the report stated.

The declines were the steepest since a similar rout in late July. The KOSPI has entered its longest losing streak in more than two years and fallen about a quarter from its July peak, while China's technology-focused STAR Market has shed more than a third of its value from its July highs.

The pressure has extended to US semiconductor stocks. The Philadelphia Semiconductor Index fell 2.7% on Monday, while Samsung Electronics and SK Hynix dropped 10% and 8%, respectively, in Seoul.

Rising bond yields put AI valuations under pressure
Higher borrowing costs are emerging as a significant challenge for a sector whose expansion depends on enormous investments in data centres, advanced chips and computing capacity.

Damien Boey, portfolio strategist at Wilson Asset Management in Sydney, told Reuters that rising bond yields and increased volatility in the bond market were beginning to influence asset prices across financial markets.

The shift in sentiment reflects concerns that higher financing costs could slow the pace of AI capital expenditure and expose companies that have committed substantial resources to infrastructure expansion.

The growing caution was also evident in BlackRock's investment outlook. The research arm of the world's largest asset manager downgraded its view on US equities to neutral on Monday, citing potential volatility surrounding the AI buildout, while turning bullish on short-term US Treasuries.

The International Monetary Fund also warned on Tuesday that a correction in AI-related investments could have global consequences. In its latest financial stability report, the IMF acknowledged AI's contribution to economic growth but highlighted increasing evidence that investment in the technology and its eventual economic benefits might fall short of expectations.

Also Read | Global Market: Japan bond yields diverge as investors weigh inflation, BOJ outlook


AI funding faces tougher investor scrutiny
Signs of caution are becoming increasingly visible in capital markets, where companies are finding that investor appetite for AI-related financing is no longer assured.

Firmus, an Australian data centre operator backed by Nvidia, shelved its planned $5 billion initial public offering last week after investor interest weakened. The decision underscored how quickly concerns about risk and valuations can disrupt funding plans in a sector that has attracted substantial capital.

Tencent Holdings' dollar and offshore yuan bond sale in Hong Kong on Tuesday also drew attention as investors assessed borrowing costs for major technology companies. Guidance for its 10-year dollar notes was around 1.15 percentage points above US Treasury yields, which stood at approximately 5.2%.

Meanwhile, SoftBank, a major technology investor, sold a record-sized high-yield debt offering last month, promising investors returns of as much as 9.75% on $4.5 billion of bonds with a 7.5-year maturity.

The borrowing costs highlight the financial pressure facing companies seeking to fund ambitious technology investments as bond yields rise and lenders demand greater compensation for risk.

The issue has also featured prominently at major financial industry conferences in Singapore. Broadcom's chief executive said on Monday that access to capital was the biggest constraint on growth, the report stated.

Investors shift focus from AI growth to returns
Despite the recent losses, investors have not abandoned the AI trade altogether. Meta Platforms shares rose 3.4% on Monday after the company released a new AI model, demonstrating that positive developments in the sector can still attract buyers.

However, semiconductor stocks are increasingly being viewed as an early indicator of whether the AI investment boom is entering a more difficult phase.

Marc Velan, head of investments at Lucerne Asset Management in Singapore, told Reuters that the sharp decline in South Korean shares reflected investor positioning and the rising cost of capital more than a fundamental deterioration in AI demand.

Velan said markets were entering a phase in which investors would pay closer attention to the commercialisation of AI technology and the returns companies generate on their investments.

That shift could distinguish businesses that can convert heavy AI spending into sustainable revenue and profits from those whose expansion depends on continued access to inexpensive financing.

The immediate question for markets is whether the latest selloff represents a temporary correction after a powerful rally or the beginning of a broader reassessment of AI valuations.

As borrowing costs rise and investors become more selective, companies across the AI ecosystem may face greater pressure to demonstrate that their infrastructure spending can deliver meaningful financial returns.

(Disclaimer: This article is based on inputs from agencies. These do not represent the views of The Economic Times)