Listened to Constellation Energy's 2026 Business and Earnings Outlook call (Mar 31), the first update since the Calpine close. Stock dropped 5 to 7% the next day on "guidance miss" optics ($11.00-$12.00 EPS vs. ~$11.72 consensus, no hyperscaler deal). Wrong thing to look at. What matters is what was explicitly excluded from base case.
Things from the call and 8-K not in the tape recap:
The 20% base EPS CAGR through 2029 is a floor, not a target. Trajectory ($6.65 in 2026 to $11.40-$11.90 in 2029) assumes nuclear at the PTC floor, 2% inflation, no new contracts beyond those announced, no buyback execution, and cash in 2028-2029 earning interest income. Unusual posture for a stock priced for perfection.
The PTC is an inflation instrument, not just a floor. At 2% inflation the 2031 PTC cap sits at $50.88/MWh; at 3.5% it moves to $56. Each 100 bps above the 2% base adds 100 bps to base EPS CAGR, mechanically, across the open position. A federally-backstopped CPI escalator on uncontracted nuclear output. No competitive generator has it.
The 147 million MWh open 2030 position is the quiet number. Contracted 48 million MWh of 2030 output (up from 12 million last year); the remainder is larger than every other US merchant nuclear operator combined. Each 1 GW of new nuclear contracting adds $0.40-$1.00/share to base EPS (1-3 points on the CAGR); gas adds $0.20-$0.50/share per GW.
The Eddystone-to-Crane capacity rights transfer is a tactical FERC workaround. PJM flagged that interconnection upgrades could push Crane's full deliverability into the 2030s. The FERC filing transfers capacity injection rights from Eddystone (running on DOE 202(c) orders past its retirement) to the 835 MW Crane site. The Microsoft PPA's 2027 start depends on FERC waivers from PJM contingent-facility stipulations. Watch the mechanism, not the restart date.
Capital markets reception was the real signal. In January, CEG priced $2.75B across four tranches to retire Calpine sub-IG debt, including $800M of 40-year notes at 5.875% (2066 maturity). Forty-year paper for a competitive power issuer at sub-6% is the tightest long-dated pricing in the sector. Moody's revised its downgrade threshold to 25% FFO/Debt from 30%; agencies are underwriting merged cash flows at materially looser metrics.
The market read this as a disappointment because no hyperscaler contract landed. The honest read: management built a floor, disclosed the sensitivity layer, and chose not to pull those levers into base. The optionality is the story; upside on 147 million MWh of open nuclear, a federally-backstopped inflation hedge, a tactical FERC path preserving Crane, and a balance sheet rated loose enough to pursue M&A without stressing the cap structure. Selling CEG at a discount to AI-utility peers is selling the baseline and giving away the sensitivity table.