LS Power to Acquire Natural Gas Power Plant in ERCOT from Constellation

LS Power agreed to acquire the Brazos Valley Energy Center, a 606 MW natural gas combined-cycle plant near Houston, to add dispatchable capacity as Texas power demand surges from AI, data centers, and electrification. The deal addresses a regulatory divestiture tied to Constellation’s Calpine acquisition, with expected close in Q4 2026 subject to approval. Closing pending Constellation transactions would bring LS Power’s national operating fleet to ~14,100 MW, reinforcing its ability to rapidly expand generation in ERCOT.

Analysis

This reads less like a one-off asset sale and more like a valuation marker for dispatchable power in ERCOT. When a private buyer is willing to pay up for an operating combined-cycle asset, it implies replacement value is moving faster than public-market expectations, which should lift the floor under merchant generation multiples and make greenfield projects look more attractive once interconnection and gas supply are secured.

The near-term winners are the public names with the cleanest Texas/merchant exposure — VST and NRG — because they benefit from any widening in forward spark spreads and from optionality on data-center load growth. A quieter beneficiary is HLI: if this is the start of a longer asset-rotation cycle, advisory fees in power/M&A can compound with relatively little balance-sheet risk. The losers are regulated utilities and power-intensive users that still rely on the market narrative of cheap, abundant capacity; if ERCOT scarcity persists, their cost of service or load-retention assumptions get less benign.

The key risk is that this is a portfolio optimization trade, not a demand supercycle confirmation. If ERCOT 2027-2028 forwards flatten, or if Texas market rules cap scarcity rents after any reliability event, the equity read-through can reverse quickly; that would show up first in merchant generator underperformance versus XLU, then in turbine/grid backlog assumptions over 6-18 months. The contrarian view is that the market may be overpricing the AI/data-center thesis: the faster these plants are absorbed into the fleet, the more likely competition erodes returns unless fuel and transmission are constrained.

Key Decisions for Investors

• Long VST vs. short XLU for a 1-3 month relative-value expression on ERCOT scarcity/merchant power; thesis works if Texas forward power prices and heat-rate spreads keep firming. Falsify on a material rollback in ERCOT forwards or a regulatory move that compresses scarcity pricing.
• Add to HLI on M&A-cycle optionality, but only on pullbacks; this is a lower-beta way to play continued consolidation in power and infrastructure. Risk/reward is attractive if transaction volume keeps inflecting over the next 2-4 quarters; thesis weakens if deal flow stalls or spreads widen across the sector.
• Watchlist, not immediate trade: GEV and ETN on any dip, as the longer-dated implication is more gas turbine, switchgear, and grid hardware demand if replacement value for dispatchable capacity keeps rising. Need confirmation from order intake/backlog before getting aggressive.
• If looking for a cleaner hedge, pair long merchant generation with a short in a regulated utility basket; the trade should work if ERCOT scarcity persists but will fail quickly if Texas power policy shifts toward intervention.