CreditSights' Cisar Sees Cracks in AI Credit Boom

CreditSights global head of strategy Winnie Cisar says the surge in AI financing is making credit analysis increasingly difficult. Investors are weighing uncertain returns, potential data center overbuilding and more complex financing structures; the report gives no financing totals or market reaction.

Analysis

Credit transmission: The key risk is not AI demand itself but whether project cash flows, power availability and tenant commitments match debt maturities. More bespoke financing can obscure where construction, utilization and refinancing risk ultimately sit; if lenders demand tighter covenants or higher spreads, marginal data-center projects may be delayed before the largest cloud platforms change spending plans. That would shift pressure toward developers and equipment suppliers with order books dependent on those projects, while better-funded platforms could gain share.

Timing and reversal: Near term, this is a diligence and risk-premium issue, not evidence of realized defaults. Over the next 1–3 months, watch issuer disclosures and financing documents for tenant concentration, take-or-pay protections, completion guarantees, debt recourse and funding costs. Over 6–18 months, commissioning delays or low utilization could expose mismatches between contracted revenue and debt service. The concern would ease if projects show timely delivery, durable contracted demand and financing that leaves meaningful equity beneath lenders.

Contrarian view: The market may conflate opaque structures with bad credit. Long-dated, creditworthy tenant commitments can support leverage; conversely, headline capex or announced capacity is not proof of bankable cash flow. With no issuer-level terms or spread data here, broad short exposure is premature.

Key Decisions for Investors

• Do not make a sector-wide short from this commentary alone; it provides no evidence of deteriorating issuer credit or mispriced spreads.
• For credit portfolios, favor transparent, senior exposure to established investment-grade borrowers over opaque, highly levered project or private-credit structures unless the latter offer demonstrably adequate covenant protection and spread compensation.
• Put data-center and AI-infrastructure financings on watch: verify debt recourse, lease/take-or-pay terms, tenant concentration, power access, completion guarantees, maturity schedule and refinancing cost before adding exposure.
• Reassess toward a relative-value short in the most leveraged project debt only if spreads fail to compensate for weak protections or disclosures show delays, cancellations, or utilization below debt-service needs; reverse that view if delivery and contracted cash flows are demonstrated.