尽管谷歌、英特尔、亚马逊和Meta等主要科技公司股价近期大幅下跌，但超大规模云服务商仍在继续对人工智能基础设施进行前所未有的资本支出。
Big Tech AI Spending Relies on Bond Market Runway Amid Stock Sell-Off Despite recent sharp stock declines across major technology firms such as Google, Intel, Amazon, and Meta, hyperscalers continue unprecedented capital expenditures into artificial intelligence infrastructure.
据摩根大通统计，自2023年以来AI支出已达1.6万亿美元，预计未来十年每年将达到1万亿美元。
According to JPMorgan, AI spending has reached $1.6 trillion since 2023 and is projected to hit $1 trillion annually over the next decade.
债务融资支撑着这一扩张，高盛估计超大规模云服商今年将发行约2500亿美元债券，总债务能力估计达5100亿美元。
Debt financing underpins this expansion, with Goldman Sachs estimating hyperscalers will issue approximately $250 billion in bonds this year, supported by an estimated $510 billion in total debt capacity.
虽然超大规模云服务商受益于强劲的云收入和合同积压，但CoreWeave等较小运营商面临越来越大的压力，其债券收益率已攀升至15%。
While hyperscalers benefit from strong cloud revenue and contract backlogs, smaller operators like CoreWeave face mounting pressure with bond yields climbing to 15%.
分析师指出，主要的长期风险在于如果AI盈利仍然遥遥无期，债券市场是否会维持低融资成本。
Analysts note the primary long-term risk lies in whether the bond market will maintain low borrowing costs if AI profitability remains distant.