Back to News
Market Impact: 0.22

GE Vernova vs. Vistra: One AI Power Stock Has Absolutely Crushed the Other, And Could Continue Doing So

+1
Energy Markets & PricesCapital Returns (Dividends / Buybacks)Credit & Bond MarketsCompany FundamentalsCorporate EarningsRegulation & LegislationTechnology & InnovationArtificial Intelligence
GE Vernova vs. Vistra: One AI Power Stock Has Absolutely Crushed the Other, And Could Continue Doing So

The article pits GE Vernova vs. Vistra for 2026, highlighting GE Vernova’s stronger financials and growth: FY2025 revenue rose 8.9% to $38.1B and net income reached $4.9B with net margin improving to ~12.8% (from 4.4%), alongside FCF of ~$3.7B and near-zero debt-to-equity. Vistra’s FY2025 results were weaker, with revenue down 12.4% to $17B and net margin falling to 5.6% (from 13.7%), and FCF of ~$129M, plus higher leverage (debt-to-equity ~4x) and a pending $4B Cogentrix acquisition facing regulatory scrutiny. Valuation also skews more favorably to Vistra (Forward P/E 16.6x vs. GE Vernova 41.0x), but the bullish argument centers on AI-driven power demand—GE’s turbine backlog is cited at $263B for FY2026 Q1 with customer upfront payments through 2030.

Analysis

Relative value favors GEV over VST because the former is monetizing an infrastructure bottleneck, while the latter is still a leveraged claim on power-price volatility. In practice that means GEV’s earnings visibility should support a higher terminal multiple if booking conversion stays tight, whereas VST’s equity can re-rate only if regional forwards rise faster than hedge costs and fuel input inflation. Over the next 1-3 months, the key catalyst is whether AI-related load actually keeps pushing customers toward behind-the-meter solutions and whether GEV can protect pricing as lead times extend. The risk is that the market is already capitalizing several years of growth; any slip in order quality, supply-chain execution, or turbine margin can compress the stock quickly even if backlog remains large. For VST, the upside is sharper but more binary: a spike in ERCOT/PJM basis or a favorable regulatory outcome could force a fast rerating, while flat power curves would leave the balance sheet exposed. The contrarian miss is that consensus is treating 'AI power demand' as one trade, when the winners split into equipment suppliers and dispatchable generators. GEV likely captures the cleaner structural growth, but VST could be the bigger convexity trade if power scarcity becomes chronic rather than episodic. The falsifiers are simple: a slowdown in GEV bookings/lead times, or a failure of regional power prices to stay elevated into the next 2-3 reporting cycles.