Key Takeaways
- • The OECD estimates AI-related deals accounted for 34% of global private credit transactions in 2025, according to figures cited in a Reuters ROI column by Altaf Kassam of Reference Portfolio Partners.
- • The Bank of England expects roughly $700 billion of data-center spending over 2026-28 to be financed through private credit, a market the Financial Stability Board sized at $1.5-2 trillion at the end of 2024.
- • Quarterly model-based valuations of illiquid AI-linked loans can smooth reported prices and make portfolios appear steadier than the underlying collateral warrants.
- • JPMorgan Asset Management estimates buildings make up only about 30% of an AI data center's cost, with chips and equipment at 70%, so depreciation of hardware poses collateral risk.
- • The FSB warns of a "first-mover incentive" in semi-liquid funds, where early redeemers can exit at stale marks and leave later write-downs to remaining investors; the FSB estimates retail investors held about 13% of US private credit in 2024.
Summary
Big Tech's AI data-center buildout is increasingly financed by private credit, according to a Reuters ROI column by Altaf Kassam of Reference Portfolio Partners. Kassam argues that AI-linked private lending is now large enough to shape how investors experience risk, not only returns. The OECD estimates AI-related deals made up 34% of global private credit transactions in 2025, and the Bank of England expects about $700 billion of data-center spending over 2026-28 to be financed this way. The Financial Stability Board sized the overall private credit market at $1.5-2 trillion at the end of 2024.
Because these loans do not trade daily, managers typically value them quarterly using cash-flow models and comparable transactions. Kassam notes this can smooth reported prices, making portfolios look steadier than the underlying businesses. Collateral is a central concern: JPMorgan Asset Management says the building accounts for only about 30% of an AI data center's cost, while chips and equipment make up 70%. If that equipment loses value faster than assumed, stale valuation assumptions could overstate how well the debt is covered.
The issue is most acute in semi-liquid funds, which offer periodic redemptions at a stated net asset value. The FSB warns of a "first-mover incentive": if a model marks a loan at 100 cents on the dollar but actual sale prices are lower, early redeemers can exit at the higher mark while later write-downs fall on remaining investors. When redemption requests spike, funds can cap withdrawals.
Retail exposure adds to the concern. The FSB estimates retail investors held about 13% of US private credit in 2024, and such exposure can reach everyday savers through default retirement options such as target-date funds. The risk is one of timing: if a semi-liquid fund pays out at last quarter's mark while loans would fetch less in an actual sale, the value gap shifts to investors who stay. Savers may only recognize the full impact when marks are adjusted or when they attempt to withdraw.
Because these loans do not trade daily, managers typically value them quarterly using cash-flow models and comparable transactions. Kassam notes this can smooth reported prices, making portfolios look steadier than the underlying businesses. Collateral is a central concern: JPMorgan Asset Management says the building accounts for only about 30% of an AI data center's cost, while chips and equipment make up 70%. If that equipment loses value faster than assumed, stale valuation assumptions could overstate how well the debt is covered.
The issue is most acute in semi-liquid funds, which offer periodic redemptions at a stated net asset value. The FSB warns of a "first-mover incentive": if a model marks a loan at 100 cents on the dollar but actual sale prices are lower, early redeemers can exit at the higher mark while later write-downs fall on remaining investors. When redemption requests spike, funds can cap withdrawals.
Retail exposure adds to the concern. The FSB estimates retail investors held about 13% of US private credit in 2024, and such exposure can reach everyday savers through default retirement options such as target-date funds. The risk is one of timing: if a semi-liquid fund pays out at last quarter's mark while loans would fetch less in an actual sale, the value gap shifts to investors who stay. Savers may only recognize the full impact when marks are adjusted or when they attempt to withdraw.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.