AI Buildout is Boom 'Not Bubble,' Says Jared Gross

Credit markets are described as increasingly jittery about additional mega-debt deals to finance the AI arms race, amid an already financing-heavy market. JPMorgan Asset Management’s Jared Gross says he views the AI buildout as a boom rather than a bubble; the article gives no deal amounts or market moves.

Analysis

The key market question is not whether AI adoption is real, but whether cash flows arrive soon enough to service the financing required to build capacity. If debt-funded data-center and power investment outruns utilization and monetization, the first vulnerability is likely issuer credit quality and refinancing flexibility—not necessarily the long-run AI thesis. That creates dispersion: infrastructure suppliers may benefit near term, while highly levered builders and capital-intensive operators could face spread widening if project returns lag funding costs. Power equipment, grid services, and data-center capacity are potential beneficiaries, but the article provides no issuer-level exposure or deal terms to establish relative winners.

Jared Gross’s view is a useful counterweight, not independent proof of returns. The immediate catalyst is the terms and reception of new large financings; over 1–3 months, watch issuance concessions, order-book quality, and credit spreads. Over 6–18 months, utilization, customer commitments, and realized returns on invested capital will determine whether capacity supports earnings or becomes stranded. A broad risk-off move could pressure even sound issuers; conversely, strong contracted demand and improving unit economics would weaken the credit-bear thesis. The contrarian risk is treating all AI capex as speculative: debt concerns could be overgeneralized across borrowers with materially different balance sheets and contracted cash flows.

Key Decisions for Investors

• Avoid a blanket short of AI equities based on debt headlines; the article supplies no issuer-specific leverage, financing terms, or cash-flow evidence. Require borrower-level diligence before taking directional exposure.
• Watch new mega-deal pricing and secondary spreads. If concessions widen and comparable AI-linked credit underperforms broader investment-grade credit, consider a relative-value underweight in the affected issuer or sector exposure versus higher-quality credit; size only after identifying the actual borrower and maturity profile.
• For the next 1–3 months, track deal demand, covenant and collateral protections, and whether financing is supported by contracted customer revenue. Weak demand or repeated repricing would strengthen the credit-caution thesis; robust books without concession would argue against it.
• Set a 6–18 month falsification test: sustained utilization, customer commitments, and returns sufficient to cover funding costs would undermine the stranded-capacity concern. Until those data are available, treat the boom-versus-bubble claim as a thesis to monitor, not a standalone trade.