Fed Financial Stability Report: 87% of Respondents List AI-Related Risks as Important Risk in the Next 12 to 18 Months
The Federal Reserve's latest Financial Stability Report shows that 87% of respondents listed AI-related risks as an important risk to U.S. financial stability over the next 12 to 18 months, a marked increase from 50% in the survey conducted in May this year; 52% of respondents listed a sharp correction in the prices of AI-related risk assets as an important risk. The report notes that as of October 2026, debt issuance related to advanced computing has exceeded $680 billion, roughly five times the $130 billion for all of 2023; data center construction also involves private credit, structured credit and project financing, and a circular financing relationship has formed between large technology companies and GPU manufacturers. The Federal Reserve estimates that forward leases and backstop guarantees currently used for data center construction have reached about $3 trillion.
Body · AI translation
Recently, the Federal Reserve released a new round of its Financial Stability Report, showing that through a survey it found that AI-related risks became the potential risk to U.S. financial stability over the next 12 to 18 months most frequently cited by respondents, with 87% of respondents listing it as an important risk, a marked increase from 50% in the survey conducted in May this year. In the survey, a sharp correction in the prices of AI-related risk assets was also the second item of concern to respondents, with 52% of respondents listing it as an important risk. The survey was conducted from September to October this year, earlier than the personnel turmoil at chipmaker NVIDIA, after NVIDIA Chief Executive Officer Jensen Huang announced his resignation on November 12. In addition, the survey also predates the U.S. midterm elections, in which the Democratic Party won control of both chambers of Congress, and the market is concerned that the Democratic Party's strengthening of regulation of the AI industry could affect investment by technology companies and data center construction. The opinions previously given by respondents did not take these factors into account. The attention to AI-related risks is closely related to the rapid expansion of investment in data centers and advanced computing infrastructure. The Federal Reserve report shows that as of October 2026, the scale of debt issuance related to advanced computing has exceeded $680 billion, about five times the $130 billion for all of 2023. The financing methods are also changing. The Federal Reserve pointed out that in addition to traditional corporate bonds, data center construction also involves private credit, structured credit and project financing. Large technology companies usually sign long-term leases for data center capacity with third-party data center developers, which then use these leases to finance data center construction. GPU manufacturers also often support advanced computing technology companies or large language model (LLM) companies in purchasing GPUs by providing residual value guarantees for the GPUs used as collateral. In addition, large technology companies and GPU manufacturers also participate in related financing arrangements through equity investments, forming a "circular financing" relationship. The Federal Reserve stated that under this "circular financing" model, large technology companies may need to bear related losses during an economic downturn, thereby to some extent buffering the transmission of risk to the entire financial system. But the report also warned that the risks these companies actually bear may be far higher than the liabilities shown on their balance sheets. The Federal Reserve estimates that the current scale of forward leases and backstop guarantees used for data center construction has reached about $3 trillion. In addition to financing risks, uncertainty about the returns on AI investment may also affect financial markets. The Federal Reserve stated that if corporate earnings unexpectedly decline, it could trigger asset price corrections, tighter credit conditions and reduced investment, thereby weakening aggregate economic demand. At the same time, the relatively high level of leverage in financial markets may amplify the impact of asset price adjustments. The Federal Reserve pointed out that hedge fund leverage, especially that of the largest hedge funds, remains at the highest level since the Federal Reserve began collecting relevant data in 2013. If the prices of AI-related assets fall sharply, highly leveraged investors may face greater losses, and the related asset selling may also further intensify market volatility and put pressure on financial intermediaries.