``` The global banking industry is "overwhelmed" and seeks to reduce "data center exposure." ```

``` The global banking industry is "overwhelmed" and seeks to reduce "data center exposure." ```

```

The surge in AI infrastructure construction is pushing the world’s largest banks to the limits of their financing capacity.

On May 3, according to the UK’s Financial Times, as data center debt levels soar, major lenders including JPMorgan Chase, Morgan Stanley, and Sumitomo Mitsui Banking Corporation (SMBC) are actively seeking to spread related risks to a broader group of investors to free up balance sheet space and maintain their ongoing lending capability.

According to reports, insiders revealed that banks are exploring private sales of debt shares and using so-called “Significant Risk Transfer” (SRT) tools to reduce their exposure to any single large borrower. Matthew Moniot, co-head of credit risk sharing at Man Group, bluntly stated: “The scale we’re talking about... far exceeds anything we’ve previously imagined. Banks will soon be overwhelmed.”

Analysts say, this pressure reflects the unprecedented scale of financing demand in the AI industry. Data center operators Oracle and CoreWeave have borrowed hundreds of billions of dollars to build facilities for AI labs across the US. This wave is not only testing banks’ risk tolerance, but also reshaping the entire structure of the data center financing market.

$38 Billion Debt Hard to Digest, Banks Seek Discounted Sales

Insiders revealed that lenders including JPMorgan Chase and Mitsubishi UFJ Financial Group (MUFG) have spent over six months trying to distribute $38 billion in construction debt related to Oracle’s data center projects in Texas and Wisconsin.

Some banks are even seeking to sell these Oracle-related loans at a discount to non-bank lending institutions. This case clearly highlights the current predicament faced by the banking sector: The scale of single project financing now exceeds the capacity of traditional syndicated loan models.

Faced with overly concentrated risk exposure, banks have recently begun to tentatively pitch various structured solutions, including SRT variants, to investors.

Previously, SRT (Significant Risk Transfer) was widely used in European banking to reduce capital requirements—banks transferred the loss risk in part of their loan portfolios to private credit funds, insurance companies, and other investors in exchange for corresponding returns. In recent years, North American banks have also adopted this tool more frequently.

However, SRTs related to data centers differ significantly from traditional models. Traditional SRTs typically involve dozens of loans, whereas the approach banks are now exploring is to carve up a single, massive and highly concentrated data center loan and transfer its riskiest portions off balance sheet.

David Lucking of the law firm Linklaters said he has seen transactions supported by a single borrower as large as $500 million; “SRT investors want to ensure the bank retains some exposure to align interests.”

Cheyne Capital portfolio manager Frank Benhamou pointed out that compared to traditional SRTs, data center transactions “have a limited number of operators, extremely high concentration, and significant construction risk,” so investors demand higher yield premiums.

Dual Pressure of Risk Limits and Regulatory Constraints

The fundamental reason banks are eager to spread risk is internal risk limits—these caps restrict the maximum exposure to a single borrower or specific industry. Once the ceiling is reached, banks can no longer finance new projects, missing out on business opportunities brought by AI infrastructure construction.

Matthew Moniot stated:

“If I were a bank’s chief risk officer, facing bankers’ requests for multi-billion-dollar credit lines for single projects, I would ask how they plan to distribute those risks.”

Meanwhile, data centers face growing public opposition in some regions of the US, adding further project risk. In April this year, Maine in the Northeast US passed a statewide data center ban, introducing new uncertainties for related financing.

Risk Warning and DisclaimerThe market carries risks, and investment needs to be cautious. This article does not constitute personal investment advice, nor does it take into account the unique investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article suit their particular circumstances. Investing based on this information is at your own risk. ```