Moody's: Private Credit Faces 2028 Test

Private credit is entering a pivotal phase as AI infrastructure spending creates new lending opportunities, but software-focused portfolios are showing signs of stress. Moody's Marc Pinto warned that the real test arrives in 2028, when a refinancing wave could reveal which borrowers can survive. The outlook is cautious for private credit managers, with credit selection and refinancing risk likely to matter more over the next few years.

Analysis

The key second-order effect is not simply that private credit is growing, but that underwriting standards are likely bifurcating by end-market. AI infrastructure loans should see tighter covenants and stronger collateralization because asset lives are shorter and cash flows are more visible, while software-heavy borrowers face a longer-duration reset risk as growth assumptions compress and exit markets remain weak. That creates a cleaner distinction between lenders with disciplined project-finance style structures and those still exposed to sponsor-friendly, covenant-lite lending.

The 2028 refinancing wall is the real catalyst, but the market will start pricing it earlier if rates stay restrictive and EBITDA doesn’t reaccelerate. The vulnerable cohort is levered software and services names that depended on cheap PIK-heavy capital to bridge the gap to profitability; those credits may not fail immediately, but they can become amend-and-extend candidates, which quietly erodes lender returns before any headline default cycle appears. A forced repricing in secondary loan markets is likely 12-24 months ahead of the maturity wall if public comparables remain soft.

A less obvious winner is the capital stack around AI buildout: data center REITs, power infrastructure, and equipment vendors can benefit from lenders competing to finance the same secular demand. That said, the opportunity is crowded and duration-sensitive—if AI capex pauses or hyperscalers delay deployment, the growth narrative can decelerate fast, leaving lenders with concentrated, large-ticket exposures. In contrast, the software credit pocket has asymmetric downside because refinancing is a balance-sheet event, not a product-cycle event.

Consensus is underestimating how much of the “private credit” label is actually becoming two different businesses: structured asset-backed lending tied to tangible infrastructure versus corporate leverage lending tied to sponsor exit assumptions. The former should deserve tighter spreads but lower loss rates; the latter is where spread compression may prove false comfort. The best risk/reward is to position for dispersion, not broad direction.

Key Decisions for Investors

• Short a basket of highly levered software names with near-term maturities vs long AI infrastructure beneficiaries (e.g., GOOGL/MSFT via equity proxies, or data center REITs like EQIX/COR if preferred) over 6-12 months; thesis is refinancing dispersion, not macro beta.
• Add selective longs in data center/power infrastructure beneficiaries on pullbacks over the next 1-3 months, but only where valuations are not already pricing flawless AI capex; target names with contracted cash flows and lower refinancing risk.
• Reduce exposure to private credit managers/heavy direct-lending platforms with outsized software exposure; the risk/reward deteriorates into 2027 as amend-and-extend activity increases and marks can lag fundamentals.
• Buy out-of-the-money downside protection on levered software credit proxies or high-yield ETFs with technology weightings for 12-18 months; downside is convex if refinancing markets tighten before the 2028 wall is addressed.
• Favor lenders with asset-backed or infrastructure-linked originations over generalist direct lenders in any pair trade; the spread between the two should widen as the market differentiates tangible collateral from sponsor-driven cash-flow lending.