Strategist Timothy Moe says hyperscaler spending of $800bn this year and an estimated $1.2tn in 2027 point to a durable AI infrastructure cycle, despite rising borrowing costs.
Goldman Sachs strategist Timothy Moe says artificial intelligence shares still have room to run, even as higher government bond yields make some investors more cautious. Speaking to Bloomberg TV, he said he remains in the “stronger-for-longer” camp on AI, arguing that capital spending by hyperscale cloud groups is still the clearest demand signal for Asia’s hardware suppliers. He pointed to estimated hyperscaler investment of about $800 billion this year and roughly $1.2 trillion in 2027 as evidence that the build-out is not easing.
That view fits a broader Wall Street narrative in which AI spending is increasingly seen as a multi-year infrastructure cycle rather than a short-lived trade. Earlier this year, Goldman said Oracle’s $25 billion bond sale had helped open the door to more borrowing for data-centre construction, while bankers at the firm have described financing AI data centres and power networks as a major opportunity. At the same time, Goldman and JPMorgan have already created ways for investors to adjust exposure to technology debt more quickly, including baskets linked to AI-related high-yield issuers.
The bullish case has not removed the macro risks. Bloomberg reported this month that rising long-term Treasury yields are adding pressure to markets even as investors keep pushing major technology shares to record levels. Higher borrowing costs can weigh on valuations and make the economics of large infrastructure programmes harder to justify, particularly if debt issuance accelerates to fund the next phase of AI expansion.
Still, Goldman has continued to frame artificial intelligence as a support for equity earnings and, by extension, for parts of the market tied to the theme. In previous research, its strategists said AI adoption and resilient growth could lift S&P 500 profits next year, while another Goldman strategist has argued investors should prefer hyperscalers to chipmakers because the market has already priced in scepticism about returns. Moe’s comments suggest that, for now, the bank sees the scale of spending as more important than the strain of higher yields.
Disclaimer: This content is intended for informational purposes only. Readers are advised to exercise their own judgement, conduct due diligence, or consult a qualified expert before acting on any information provided.
