Fed Financial Stability Report: AI enthusiasm pushes up stock valuations, advanced computing-related debt issuance exceeds $680 billion
The Federal Reserve released its second-half 2026 Financial Stability Report, stating that stock valuations remain persistently elevated, driven in part by market enthusiasm for artificial intelligence. As of October 2026, debt issuance related to advanced computing had exceeded $680 billion, roughly a fivefold increase from $130 billion for all of 2023, with outstanding public debt estimated at more than $1.1 trillion. The report said 87% of surveyed market participants listed artificial intelligence as a significant risk to the financial system, a sharp jump from 50% in the spring survey.
Why read this
AI score 69
The report provides specific figures showing advanced computing-related debt rising from $130 billion to more than $680 billion, offering a clear view of the changes in the financing structure behind the AI investment boom.
Topics
Body · AI translation
Article body
On Friday, US Eastern Time, the Federal Reserve published its 2026 second-half Financial Stability Report (the "Report") on its official website, summarizing the framework of the recent resilience of the US financial system from aspects including asset valuations and debt levels. The Fed believes that current stock valuations remain persistently elevated, driven in part by market enthusiasm for "artificial intelligence," while at the same time the scale of US AI-related debt is expanding rapidly.
The Report shows that the risks brought by artificial intelligence and the rapid growth of investment in this field have become respondents' foremost concern for financial stability.
"Asset valuation pressures remain persistently elevated. Equity valuation metrics have remained high, partly due to market fervor over 'superintelligence,'" the Fed added, noting that current market fervor has cooled somewhat amid concerns over corporate capital expenditure and debt issuance plans.
The Report focuses on debt financing related to advanced computing. According to the Report, as of October 2026, debt issuance related to advanced computing had exceeded $680 billion, roughly a fivefold increase from $130 billion for all of 2023; investment-grade technology companies issued most of this debt, and advanced computing-related borrowing rose from less than 6% of nonfinancial investment-grade corporate bond issuance in 2023 to about 31% in 2026. As of October 2026, outstanding public debt related to advanced computing is estimated at more than $1.1 trillion.
The Report states that cash outflows corresponding to capital expenditure indicate that the expansion pace of this investment cycle may soon exceed the telecommunications investment boom of the 1990s.
The debt issuance structure is also changing: from being dominated by corporate bonds to a model combining public debt, private debt and structured credit.
The Report states that long-term contracts, leases and guarantees all constitute channels through which hyperscalers and other large enterprises with higher credit ratings support financing for projects developed by enterprises with weaker credit quality, without having an immediate, direct impact on the balance sheets of the stronger-credit enterprises. For example, hyperscalers typically sign long-term lease contracts with third-party data center developers for data center capacity, and the third-party developers then use such lease contracts as collateral to obtain financing for data center construction. In addition, graphics processing unit (GPU) manufacturers typically provide guaranteed residual values for GPU collateral to support advanced computing technology companies or large language model (LLM) companies in purchasing their GPUs. Such debt financing, which connects hyperscalers and GPU designers with smaller enterprises, is often accompanied by equity investments from hyperscalers and GPU designers, thereby creating a "circular financing" dynamic.
This structure means that once the AI industry enters a downturn, the scale of related debt defaults may be higher than what current corporate balance sheets reflect, because these large technology companies also bear substantial future lease commitments, external guarantees and other complex obligations.
Regarding financial sector leverage, the Report states that hedge fund leverage has reached its highest level since data began in 2013, and is concentrated in a limited number of small funds, broadly supporting multiple trading strategies including Treasuries, interest rate derivatives and equities. The banking sector remains sound and resilient overall. Banks continue to report regulatory capital levels near historical highs, but fair value losses on fixed-rate assets remain large and sensitive to changes in long-term interest rates.
The survey of market participants cited in the Report shows that 87% of respondents listed artificial intelligence as a "significant risk" to the financial system, a sharp jump from 50% in the spring survey, far exceeding the 52% for a pullback in risk assets and 48% for rising long-term interest rates. Respondents also warned that AI advances are lowering the threshold for cyberattacks, with cyberattacks and fiscal debt sustainability listed as the risk categories with the most severe potential consequences.
(Source: Shanghai Securities News) Middle of the article Embedded advertisement