The Energy Stock to Watch in October Is Not Another Oil Major

Every October, cable treats “energy” like one green bar on the sector board. Oil jumps, XLE follows, and the story ends at the pump. That camera angle is getting lazier by the week. My take heading into the month is simple. The sector still has a real commodity tape with crude still firm. The better multi-year watch is the company selling scarce, always-on power into the AI buildout. That name is Constellation Energy (CEG), the country’s largest commercial nuclear operator. It is a different bet than another integrated oil major priced like a crude coupon. ## Energy is two businesses right now Oil still matters. West Texas Intermediate has been trading in the mid-$90s and Brent near $100 as inventories stay tight and Middle East politics refuse to go quiet. U.S. producers and the big integrated names can still throw off cash when barrels hold, and diesel still shows up in freight bills for ordinary America. That trade is real. It is also cyclical. Peace talk headlines, OPEC decisions, and weaker demand can reprice it in a week. The second business is electricity. Data centers and AI loads are pulling power demand higher after two decades of almost flat U.S. growth. Research shops keep mapping multi-year jumps in data-center load, and the grid cannot invent baseload overnight. Gas turbines help, which is why equipment order books run years deep. Nuclear helps more when a customer wants carbon-free power that runs around the clock without waiting on weather. Oil stocks ride the barrel. Power stocks ride signed megawatts and plant scarcity. October is when those two stories get confused again. ## Why Constellation sits at the choke Constellation already owns what hyperscalers cannot build in a quarter: licensed nuclear plants that run most hours of the year. After closing the Calpine deal early in 2026, the company is no longer a pure nuclear story. It added a large gas and geothermal platform and became one of the biggest competitive generators in the country. Nuclear is still the crown jewel for AI buyers who want firm, clean power they can lock under long contracts. The contract book is the part Wall Street still under-tells. Meta locked a 20-year deal around about 1,121 megawatts tied to Constellation’s Clinton plant in Illinois, starting mid-2027. Microsoft’s Crane restart, the old Three Mile Island Unit 1, is the other flagship nuclear restart story nearby. In the second quarter alone, Constellation said it signed about 920 megawatts of additional long-term nuclear power deals with investment-grade customers, with average terms measured in decades, not quarters. That is the mechanism. Scarce plants plus multi-year offtake turns a utility-looking ticker into a contracted cash-flow machine for AI load. Management has talked base earnings growth on a high-teens to 20% path into the back half of the decade, and it raised 2026 adjusted operating earnings guidance into the $11.50 to $12.50 range after Calpine and those new contracts showed up. Shares recently traded near the mid-$260s while a chunk of the Street still carries targets well above $300. I treat those targets as fee-driven context, not a promise. ## What October actually tests October will not flip Meta’s Clinton start date. The month tests whether investors still price Constellation as a messy power trader or as a contracted nuclear platform. Watch the next earnings wave into early November for fleet performance, how much clean output sits under long deals by 2030, and whether Calpine integration is still adding cash without turning into a pure gas bet. Oil names can still have a loud month if barrels spike again. GE Vernova remains the factory bottleneck for turbines and grid gear, and that story already got airtime as a market dark horse. Speculative small modular reactor names can still rip on headlines while sitting years from real revenue. Constellation is the boring-sounding seat that already sells what AI campuses need: firm megawatts from plants that exist. ## What would change my mind I am not writing a blank check. Nuclear outages that stretch, a restart schedule that slips hard, or a wave of hyperscaler spending cuts that freezes new long-term power deals would all hurt the story. A pure collapse in power prices without enough contracted cover would also matter. So would leverage and integration risk if Calpine stops looking like a complement and starts looking like a distraction. If those show up together, the October watchlist should shrink. Until then, the risk I care about more is investors treating every energy ticker as an oil ticker. ## Bottom Line October energy is a split screen. Crude near $95 to $100 keeps oil majors in the conversation, but the multi-year choke is AI power. Constellation Energy is the stock I would watch first because it already pairs the largest U.S. commercial nuclear fleet with long hyperscaler contracts and a bigger generation platform after Calpine. The product is scarce baseload under paper that lasts years, while oil stays a barrel coupon and turbine makers stay a factory backlog. CEG is the one selling the electricity the boom actually burns.