The Financial Stability Board (FSB) published its "Report on Vulnerabilities in Private Credit" on 6 May 2026. The report notes that private credit grew to an industry‑estimated $1.5–2.0 trillion by the end of 2024 and increasingly finances large AI‑related infrastructure spending, such as data centers and compute capacity.
Why this matters now
The FSB treats private credit as part of a broader private finance ecosystem rather than an isolated niche. An industry scenario cited in the report estimates AI‑infrastructure capex of about $2,900 billion for 2025–2028; of that, $1,500 billion could be externally financed and roughly $800 billion could take the form of private credit. The scale and timing of these investments help explain why some financing may shift to more flexibly structured, non‑bank channels.
Bank links and hidden exposures
The FSB emphasizes that private credit is often connected to banks in multiple ways: committed credit lines, portfolio financing arrangements, and other structures that can create leverage at the fund level. Member‑country data point to roughly $220 billion of drawn and undrawn bank facilities to private credit funds, yet commercial data providers estimate materially higher exposures—even multiples of that figure—indicating gaps in data and transparency. Such discrepancies can produce divergent pictures for supervisors and market participants, increasing the risk of rapid confidence erosion in a stress event.
Credit quality, leverage and delayed deterioration
Many private credit borrowers lack public ratings; where external estimates exist, ratings around "single B‑" are common and increasing leverage is noted. The FSB warns that deterioration may not show up as an immediate wave of defaults. Contractual reliefs and restructuring options in private credit structures can postpone recognition of losses, creating a delayed deterioration dynamic. The report recommends monitoring broader distress indicators (for example, selective default or distressed exchange), which may signal credit weakening earlier than headline default rates.
Valuation, liquidity and data‑quality risks
Valuation practices in private credit tend to be less frequent and more discretionary—typically quarterly—which can understate volatility in calm times and make losses appear to emerge suddenly when they are finally marked to market. Liquidity risk is also salient: while many private credit vehicles are closed‑end, an increasing number of products promise some redemption or liquidity features. If asset positions are long‑dated and illiquid, investor redemptions or tighter financing terms can quickly translate into forced sales or refinancing stress.
AI projects add another dimension: uncertain business models and revenue generation mean a single negative shock could simultaneously hit both credit quality and financing appetite.
Supervisory blind spots
The report repeatedly calls out data shortages and definitional inconsistencies. Exposures are often spread across multiple entities, leverage appears at several levels, and bank relationships are not always transparent. From a financial stability perspective, this matters because some risks only become measurable once stress mechanisms activate—when financing tightens, collateral requirements rise and losses crystallize.
Conclusion
The FSB does not argue that private credit growth is inherently abnormal, but it warns that the financing needs tied to large‑scale AI infrastructure could accelerate credit growth and strengthen channels that transmit stress across the financial system. Key vulnerabilities include multi‑layered leverage and bank interconnections, valuation and liquidity weaknesses, and gaps in data and transparency—all of which raise the chance that a negative shock could propagate more quickly and broadly than conventional stress tests suggest.
This article is not investment advice or a recommendation.
