Constellation Energy Corporation (CEG)

FY2025 adjusted EBITDA grew approximately 20%+ driven by higher merchant power prices and new corporate PPA signings at premium rates. Revenue growth far exceeds regulated utility peers and most IPP competitors. The growth is structural rather than cyclical — driven by demand, not just commodity price spikes.

Constellation operates roughly 10% of total US generation capacity and over 50% of US nuclear capacity. In the emerging data center PPA market, CEG has an outsized share given its clean baseload profile. The Three Mile Island restart adds 835 MW of carbon-free capacity that has already been contracted to Microsoft at premium terms.

Nuclear-generated clean energy is scarce and in high demand from ESG-conscious corporates and hyperscalers. Constellation is signing PPAs at $80-100/MWh — significant premiums to wholesale prices — because customers value 24/7 carbon-free attributes. This pricing power is structural as long as data center demand exceeds clean energy supply.

Constellation's 'product innovation' is primarily about contracting structure — developing behind-the-meter nuclear PPAs and clean energy certificates that match hyperscaler requirements. The TMI restart is a novel project but relies on proven technology. Actual innovation in nuclear technology (small modular reactors) is happening elsewhere.

Corporate PPA contracts typically run 10-20 years, creating long-duration revenue visibility. However, the underlying commodity (electricity) is fungible — when contracts expire, customers can theoretically switch to other clean energy providers. The switching cost is really about availability: there simply aren't enough alternative 24/7 clean sources.

No traditional network effects apply to power generation. However, Constellation's geographic concentration of nuclear assets in PJM (the highest-demand grid region) creates a locational advantage that is partially self-reinforcing — data centers locating near its plants further increases local demand.

The regulatory moat for nuclear is the widest in any industry. Building new nuclear plants takes 15-20 years and costs $10-15B+ per GW in the US. NRC licensing is extraordinarily complex. Constellation's existing fleet, with licenses extending to 2040-2060, represents assets that are literally irreplaceable within any reasonable investment horizon.

Constellation's nuclear plants are fully depreciated sunk costs generating massive cash flows. Marginal generation costs for nuclear are extremely low ($20-30/MWh), creating enormous operating leverage to higher power prices. Maintenance capex is manageable relative to revenue. The fleet generates $3-4B+ in annual free cash flow.

FY2026 earnings estimates have been revised upward 25%+ as analysts incorporate new PPA signings and higher merchant power price assumptions. The revision trend is strongly positive and accelerating. However, the base has risen so dramatically that maintaining the upward revision trajectory requires continued positive surprises.

The AI power demand narrative is powerful and has propelled CEG into the spotlight. Media coverage of the TMI restart, Microsoft deal, and nuclear renaissance is overwhelmingly positive. However, regulatory pushback on nuclear plant PPA structures (FERC concerns about grid reliability impacts) and NIMBYism around TMI restart create countervailing negative headlines.

CEO Joe Dominguez has been a compelling advocate for nuclear energy's role in the AI era, but the company is navigating uncharted territory with the TMI restart and novel PPA structures. Capital allocation is solid — buybacks and dividend increases — but the temptation to over-invest during a euphoric period is a risk to monitor. The Calpine acquisition adds complexity.