Scorecard

The post-OBBBA pipeline collapse is not a gradual slowdown but a concentrated step-down in 2028–2029. The pre-OBBBA pipeline burns down in 2026–2027; post-OBBBA replacement additions are blocked by FEOC supply chain constraints and the physical work test. Gas turbines cannot respond on the relevant timeline. The gap is already visible in PJM's first-ever capacity auction shortfall for 2027/2028.

Wrong if: PJM's 2028/2029 capacity auctions clear without reliability events AND clean capacity additions in 2028 come in above 60 GW

The White House Ratepayer Protection Pledge (March 4, 2026) cannot resolve a capacity shortfall driven by timeline physics: gas turbines have 7-8 year manufacturing backlogs, PJM interconnection adds 1.75+ years on top, and solar/battery storage — the only technology deployable in time — is being actively suppressed by policy. The shortfall is already visible in PJM's first-ever missed reliability target (2027/2028 auction, 6,623 MW short) and NERC's 'high' risk classification for PJM, MISO, and ERCOT beginning 2029. The Pledge has no enforcement mechanism, no state regulatory integration, and no interaction with FERC co-location rules.

Wrong if: Gas permitting reform produces 10+ GW of new combined-cycle capacity online in PJM by mid-2027, OR the Pledge produces a concrete FERC-codified enforcement mechanism within 90 days that demonstrably stabilizes PJM capacity prices, OR AI efficiency gains reduce 2027-2028 data center demand projections by 30%+ before the shortfall materializes

Three simultaneous features distinguish this crisis from every prior oil shock: no swing producer outside the disruption, an actuarial (not kinetic) closure mechanism with a quarter-scale reopening lag, and a storage countdown that converts logistics shock to upstream supply destruction past Day 25. The consensus $70 year-end price baseline is built on analogies that don't apply.

Wrong if: The Strait reopens to commercial shipping at scale before March 24, OR the DFC reinsurance backstop produces verified VLCC transits at scale by March 19, OR the IEA coordinated release suppresses Brent below $90 for 60 consecutive days

Capex is running years ahead of enterprise adoption. The gap gets papered over until it doesn't.

Wrong if: Enterprise AI spend accelerates past $200B annually by Q4 2025