
Key participants in AI financing seek to reduce positions! Morgan Stanley considers selling part of its exposure related to data centers

As one of the key participants in the artificial intelligence financing competition, Morgan Stanley is considering selling part of its data center-related exposure through a method known as "Significant Risk Transfer" (SRT). Morgan Stanley is also exploring other ways to hedge or syndicate some of its data center risks, and it cannot guarantee that these early-stage SRT negotiations will ultimately lead to a transaction
As a key player in the artificial intelligence financing competition, Morgan Stanley is considering selling part of its data center-related exposure through a method known as "Significant Risk Transfer" (SRT).
According to media reports citing anonymous informed sources, Morgan Stanley has initiated preliminary contacts and discussions with potential investors regarding an SRT linked to a portfolio of loans related to artificial intelligence infrastructure.
SRTs backed by data center asset exposure are still an emerging niche in the credit risk transfer market. In this market, banks hedge their credit risk, manage capital adequacy ratios, and free up balance sheet space for further lending by selling credit-linked notes to institutional investors.
Morgan Stanley is also exploring other ways to hedge or syndicate some of its data center risks, and it cannot guarantee that these early-stage SRT negotiations will ultimately lead to a transaction.
In October of this year, Morgan Stanley arranged over $27 billion in debt financing and approximately $2.5 billion in equity financing for a special purpose vehicle (SPV) related to the Hyperion data center project developed by Meta Platforms Inc. in Richland Parish, Louisiana.
Additionally, Morgan Stanley has led recent junk bond issuances for TeraWulf Inc., Cipher Mining Inc., and Applied Digital Corp., with some of the raised funds earmarked for financing new data center facilities.
Morgan Stanley's strategists predict that by 2028, spending by large cloud computing companies on data center infrastructure projects will reach approximately $3 trillion. The bank estimates that only about half of this can be covered by their own cash flow, with the majority needing to be raised through the debt market.
This surge in lending could lead to excessive concentration exposure risk for banks in a few companies.
As a company that was once known for its stability but has now borrowed hundreds of billions of dollars and deeply tied its fate to the AI boom, Oracle's cost of debt default protection has surged significantly in recent months. A research report from Morgan Stanley pointed out that banks and other lending institutions providing financing for its construction loans are likely one of the key factors driving this cost increase.
Earlier this year, Morgan Stanley promoted an SRT product linked to a portfolio of private market fund loans. According to a survey released by Bloomberg in June this year, global SRT sales are expected to grow at an annual rate of approximately 11% over the next two years
