软银考虑发行200亿美元债券以入股OpenAI,信用风险随之上升
SoftBank Credit Risk Rises As Firm Mulls $20 Billion Bond Sale For OpenAI Stake

原始链接: https://www.zerohedge.com/markets/softbank-2036-bonds-tumble-firm-mulls-20-billion-bond-sale-openai-stake

软银集团(SoftBank Group)拟发行100亿至200亿美元债券的消息传出后,其2036年到期债券价格出现下跌。此次潜在的美元及欧元债券发行,旨在为一笔400亿美元的过渡性贷款进行再融资,该贷款此前用于支持软银对OpenAI的650亿美元巨额投资,并为未来的人工智能项目提供资金。 为扩大投资者基础,据报道软银正在考虑采用144A债券发行方式——这是其十多年来未曾使用的策略——以吸引美国机构投资者。尽管该公司确认正在评估再融资方案,但强调具体的发行规模和时间尚未确定。 市场反应反映了对软银杠杆率上升及潜在供应风险的担忧,因为该公司正在加速推进其以人工智能为核心的债务策略。此举正值超大规模云厂商在极端资本支出背景下,今年已举债超过4100亿美元用于建设全球AI基础设施,从而对债券市场造成了巨大压力。

相关文章

原文

SoftBank Group's 8.5% notes due 2036 fell as much as 2.5 cents to about 98 cents on the dollar after Bloomberg reported that Masayoshi Son's conglomerate is considering a $10 billion to $20 billion bond offering, prompting traders to price in additional leverage and supply risk as the investment firm plans to refinance the massive bridge loan backing its OpenAI investment.

Sources familiar with SoftBank's plans said the potential bond offering could be denominated in dollars and euros, though they stressed that its size and timing could change. Proceeds would be used partly to repay a $40 billion bridge loan secured earlier this year to fund SoftBank's OpenAI investment and partly to finance additional artificial intelligence investments.

Masayoshi Son's junk-rated conglomerate plans to invest more than $65 billion in OpenAI by October, funding one of the largest private AI bets in history partly with borrowed money.

"We are considering various options to refinance the bridge loan, but nothing has been decided, including the amount for each," a SoftBank spokesperson told the outlet. 

The report noted:

Unlike most of its international offerings, SoftBank is exploring a 144A format for the first time in more than a decade, which would allow the notes to be sold to institutional investors in the US, the people said. That would help the company tap a larger pool of capital and potentially boost demand for the sale, they added.

US hyperscalers are expected to pour trillions of dollars into data centers and other AI infrastructure over the coming years. These firms have already borrowed more than $410 billion so far this year, putting pressure on debt markets. There are also $3.1 trillion in hyperscaler off-balance-sheet commitments that have come into question, as we noted in our latest report here.

联系我们 contact @ memedata.com