Musk Wants the Blades. Howmet Got Hit About 11%.Elon said he would cast the blades. Howmet paid the bill in one rough session. Howmet Aerospace (HWM) got hit again as the market replayed SpaceX's plan to cast industrial gas-turbine blades and vanes in Texas. Shares have already taken a multi-day beating on the idea that Musk can short-circuit the hottest bottleneck in the power build for AI data centers. One rough session alone knocked the stock about 11%. This is not a boring supplier story. It is a fight over who owns the scarce parts that keep new power plants from showing up on time. ## The Scare Is Simple Natural-gas turbines need exotic hot-section blades. Only a handful of shops can cast them at scale. Howmet holds a huge share of that niche. When Musk said in-house casting at SpaceX could pull turbines online by as much as 18 months, investors sold first and asked questions later. If the world's best-funded industrial builder can make its own blades, the shortage premium in Howmet's multiple looks less safe. GE Aerospace answered in its own way, agreeing to buy casting specialist Consolidated Precision Products for about $11.75 billion. GE is paying up to own more of the category, not running from it. The AI power story has been simple for two years. Chips need juice. Juice needs turbines. Turbines need blades. Everything else is secondary until those three links clear. ## Why the Panic Can Be Wrong Bernstein and others argue SpaceX is solving a captive power problem for its own AI build, not launching a merchant blade business tomorrow. Channel checks have put a multi-year timeline on real SpaceX product. Howmet's industrial gas-turbine work was already growing fast, and the company has been adding capacity because the queue through the end of the decade is real. Casting aerospace-grade blades is not a weekend CNC project. Qualification, yield, and metallurgy take years. Musk can throw money and engineers at the wall. History says the wall still wins for a while. Howmet also sits in commercial aerospace and defense, not only industrial gas turbines. A SpaceX foundry aimed at captive power does not erase jet-engine aftermarket demand overnight. The market is pricing the scariest version of the story first. ## What Investors Should Actually Watch Watch whether Howmet's industrial turbine growth holds when the next quarter reports. Watch whether GE's CPP deal closes clean and whether other OEMs keep locking long-dated casting supply. And watch whether SpaceX job postings and foundry buildouts turn into qualified parts, or stay as a press-cycle threat. The stock can keep whipping on every Musk sentence. A rough day on a foundry headline is still different from the franchise dying overnight. Scarcity trades die when new capacity shows up. They do not die because someone announced a foundry plan on social media. ## Bottom Line Howmet is priced like the scarce blade shop in an AI power boom. SpaceX wants to punch a hole in that scarcity for its own turbines. The selloff is the market pricing a future competitor years early. The story to own or avoid is still whether the power build stays tight long enough for Howmet's backlog and pricing power to matter more than one foundry in Bastrop.