US Steel & Aluminum

The question

Are US mini-mill steelmakers and specialty-alloy producers priced for a mid-cycle mean reversion in spreads, or for a mid-decade regime in which domestic demand for infrastructure, transmission, and aerospace alloys is structurally higher and the tariff perimeter on imports is structurally wider?

Base rates

The reference class for integrated and mini-mill producers is narrower than most investors assume. Over the last thirty years there have been three distinct regimes: 1994-2001 (over-capacity, minimal returns on capital), 2002-2008 (China-driven demand shock, sector ROIC doubled), and 2009-2019 (slow grind with cyclical spikes). The mini-mill cohort -- Nucor and Steel Dynamics in particular -- earned a sector-adjusted return roughly 2x the integrated peers across all three regimes because the scrap-and-DRI cost structure reset faster than blast-furnace costs when prices moved.

The specialty alloys reference class is different again. Titanium, nickel-superalloy, and specialty stainless producers run on aerospace build rates, not construction steel prices. Their base-rate correlation to HRC steel is below 0.3. During the last aerospace up-cycle, the specialty cohort compounded earnings roughly 18-25% annually through the middle years of the cycle.

Consensus forward earnings for the mini-mill operators embed steel spreads that mean-revert to a roughly 2015-2019 average by 2027. That is a reasonable prior for an un-interrupted global steel market. It is a less reasonable prior for a market in which Section 232 tariffs have been in place for eight years, have been widened in scope, and are now politically bipartisan.

Why the consensus view is wrong (or incomplete)

The sell-side treats the tariff perimeter as a temporary distortion that will normalize. That framing ignores three facts. First, the tariff regime has survived two administrations with opposing trade philosophies and has been broadened rather than narrowed. Second, the end-market demand mix has shifted materially toward transmission and distribution steel, data-center structural steel, and reshored-factory construction, each of which is more domestic-sourced than the historical mix. Third, the incremental cost of domestic EAF capacity has risen materially because of power-price inflation and the time required to interconnect, which raises the incentive price for new domestic supply even inside the tariff perimeter.

The consensus view is also incomplete on specialty alloys. The market still treats titanium, nickel superalloys, and specialty stainless as proxies for the commercial aerospace cycle alone. In the current cycle, the defense-industrial rebuild and the medical-device segment together contribute a larger share of incremental revenue than in prior cycles, which reduces the cyclicality of the earnings stream without meaningfully reducing the upside.

Finally, the scrap-flow question matters. If EAF capacity additions continue, prime scrap availability becomes the binding constraint. That favors the operators with integrated DRI or pig-iron supply, which is a specific and identifiable sub-set of the cohort.

Position construction

The book organizes into three sub-books.

Low-cost mini-mill core (~47%). NUE at 20% and STLD at 20% are the anchors -- both operate at the left end of the domestic cost curve, have DRI or pig-iron integration under construction or operating, and run returns-on-capital roughly double the integrated cohort across the cycle. CMC at ~6.6% completes the mini-mill exposure with rebar and long-product leverage tied directly to infrastructure and transmission.

Specialty alloys (~24%). ATI at ~11.5% and CRS at ~12.3% are the titanium and nickel-superalloy producers feeding the aerospace, defense, and medical end-markets. These positions are sized nearly as large as the mini-mill anchors because their earnings streams have lower correlation to the mini-mill book and provide genuine diversification within the sector.

Integrated, aluminum, and raw material (~29%). CLF at ~6.9% is the integrated steel exposure with an auto-sheet concentration; sized below the mini-mill anchors to reflect the structurally higher cost base. TX (~7.2%) and GGB (~7.1%) provide regional diversification into Americas long-products producers that benefit from the same USMCA-aligned demand stream. CENX (~3.2%) and KALU (~1.3%) are the primary and rolled-aluminum exposures. HCC (~3.6%) supplies metallurgical coal into the global blast-furnace cohort and is sized small to reflect commodity-price beta. ZEUS (~0.3%) is a small-cap service-center position held at a minimum weight.

Asymmetric payoff

If US infrastructure, transmission, and re-shored manufacturing investment continue along their current trajectories and aerospace build rates recover on schedule, the weighted book returns roughly 14-22% annualized over three years. If domestic demand flat-lines and steel spreads mean-revert to the 2015-2019 average, the book returns -5 to -12% with the specialty-alloy sub-book providing partial offset. If a meaningful incremental infrastructure or transmission package moves through, or if aerospace rates surprise to the upside, the right tail is 30-45%.

At 55% base, 25% bear, and 20% bull, expected value is roughly +12 to +18% annualized versus an SPY base rate near +8%. The asymmetry comes from the low-cost position of the mini-mill anchors -- their downside is bounded by conversion-cost economics, while their upside is open-ended when spreads widen.

Three things that would change our mind

1. Formal weakening of Section 232 scope or enforcement, particularly any carve-out for major low-cost exporters, that is not offset by equivalent domestic-content provisions in infrastructure or defense procurement.
2. A sustained rise in domestic prime scrap prices relative to HRC that compresses the cost advantage of EAF operators without a corresponding rise in finished-product pricing.
3. Commercial aerospace narrowbody rates slipping by more than six months with supplier-quality rather than OEM causes cited, which would signal that specialty-alloy demand is decelerating ahead of reported revenue.

What we are explicitly NOT betting on

We are not betting on a specific tariff headline or a specific infrastructure-bill pass-through number. We are not betting on higher HRC prices in isolation; the thesis is about spreads and volumes together. We are not betting on turnaround stories at integrated mills with stranded high-cost assets. We are not betting on Chinese export discipline. The thesis requires only that domestic demand for structural steel, transmission hardware, and aerospace alloys continues on its current trajectory and that the tariff perimeter remains roughly as it is today. Both are observable and slow-moving conditions, and the portfolio is sized for them.