AI Investment Boom Faces Investor Reckoning as Returns Lag Spending Spree

A major shift is underway in the artificial intelligence investment landscape, with global brokerage Jefferies warning that the current AI capital expenditure (capex) cycle is more likely to end due to impatient investors demanding actual returns, rather than tech giants simply cutting their spending. This comes as a 'massive wealth transfer' is visibly boosting North Asia market power, while leading US tech firms are increasingly relying on debt to fund their ambitious AI projects. The combined market capitalization of South Korea and Taiwan has surged from US$3.2 trillion in early 2023 to US$9.8 trillion, largely driven by their dominant semiconductor industries, highlighting where AI capital is flowing globally. This dynamic has seen major US hyperscalers like Microsoft, Alphabet, Amazon, and Meta underperform the S&P 500 since late May, raising serious questions about the sustainability of their AI spending. Many of these firms have ramped up their debt funding, issuing US$144 billion in bonds this year compared to US$83 billion in all of 2025, to finance AI infrastructure development. The Bank for International Settlements (BIS) has also echoed these concerns, cautioning that excessive investment fueled by debt, without corresponding returns, could lead to market instability and even recession, drawing parallels to past tech bubbles. Going forward, the focus will intensify on whether these massive AI investments can translate into tangible profits for US tech giants. Investors will be keenly watching for concrete Return on Investment (ROI) and sustainable business models, rather than just growth in capital expenditure. The potential for 'massive capital destruction' remains a significant risk if the promised returns fail to materialize, signaling a critical juncture for the global AI market as investor patience wears thin and the market demands accountability.