JPMorgan, Morgan Stanley, SMBC and MUFG are quietly shopping pieces of the AI debt they helped build, a sign that the lenders closest to the money have started hedging their own bet. Six months ago, banks raced to underwrite data center loans tied to Oracle, Meta and the broader AI buildout. Now some of those same banks are trying to move that debt off their books through significant risk transfers, deals that let a bank keep a loan on paper while shifting its riskiest slice to outside investors.
The banks are bumping up against internal limits on how much AI-linked exposure any one lender can hold. One case shows how stuck this has become: a group including JPMorgan and MUFG has spent more than six months trying to distribute $38 billion in construction debt tied to a Texas and Wisconsin data center leased to Oracle. Some lenders are now willing to sell pieces to non-bank buyers at a discount. Matthew Moniot of Man Group told the Financial Times the deal sizes are "out of scale to anything we've thought about, ever."
Oracle is the clearest warning sign. Its five-year credit default swaps climbed to around 151 basis points in early October, levels last seen in 2009. S&P cut the company's rating to BBB-, a single notch above junk, in July. Demand for AI-linked bonds has also thinned, with order books shrinking from roughly five times the offering size in February to under two times by July.
The broader rate environment adds pressure. The 10-year Treasury yield hit 5.304% this month, its highest since 2002. The Federal Reserve's September minutes named AI-related borrowing as a contributor, though Capital Economics points to inflation and government deficits as the bigger driver. Either way, the cost of financing a data center has risen sharply.
Source: Hacker News · Summarized by HeadlinesBriefing