Hyperscalers are committing anywhere between $600-700B in Capex for 2026. Most of this money will be spent on infrastructure build-out, specifically data-centers (DC).

    Every DC before it ever touches a GPU or rack, needs steel. It's estimated that each DC will need >20,000 tons of steel.

    This information isn't new, and SLX, a pure global steel ETF already ran up +34% in the past year, but I do think there are still meaningful bets remaining within the steel players.

    My bet is on CMC where I entered into a small position.

    Commercial Metals Company, at a $7B market cap and 11x forward P/E, produces the steel rebar and structural long products that's poured into concrete foundations, slabs and above-ground framing.

    In other words, for you to get your DC build out started, you need products from CMC before you do anything else with the DC.

    CMC ended Q3'25 with a net-income margin of 8.4%, sits on $2B in cash against total liabilities of $5B (very health) and most importantly, a backlog of contract awards on data center, energy and public works projects.

    Their play is profitable, high margin contracts and not chasing volume.

    But there is nuance.

    If you want a "safer" bet, then NUE (Nucor) is the way to go. Trading at a higher premium of 15x forward PE, It's the largest steel manufacturer in North America and supply 95% of steel products that go into a data center. So if CMC kicks off the build (e.g foundations), NUE not only starts it, but finishes the build.

    But at $36B and 5% net-income margin, I truly believe NUE is probably priced to perfection already.

    CMC doesn't cover the full data center stack the way Nucor does but what if they don't need to?

    At 8.4% net income margin (vs 5% NUE), they're winning the contracts that matter and continue to use the micro mill technology, (a facility capable of producing rebar continuously in a single process), to beat the competition on margin.

    Add to all of this:

    –we have +25% Steel tariffs that is bound to redirect construction demand domestically

    –an administration that loves to reward US manufacturers with Federal contracts.

    –nuclear and energy infra buildouts in the US and EU stimulus packages to fund infra

    CMC doesn't need a dominant share, but once it takes a bite of the pie, it should be more than enough for them to deserve a re-rating.

    Let’s talk about steel manufacturing for a moment
    byu/PositionJournal inwallstreetbets



    Posted by PositionJournal

    Leave A Reply
    Share via