Steel Dynamics, Inc. (STLD): Deep Dive Research Report

Basic Materials | Nasdaq: STLD | Research as of 20 October 2026

Reporting periods used. Steel Dynamics reports quarterly, and its fiscal year ends on 31 December. Q3 results usually come out in the third week of October, so the most recent period was due around 19-20 October 2026. It was released on 19 October 2026 after the close, and the call was held on 20 October 2026. The six results events used here are:

• Q2 2025 call, 22 July 2025
• Q3 2025 call, 21 October 2025
• Q4 2025 call, late January 2026
• Q1 2026 call, 21 April 2026
• Q2 2026 call, 21 July 2026
• Q3 2026 results release, 19 October 2026, plus the company's mid-quarter guidance release of 17 September 2026

No transcript of the 20 October 2026 call was available when this was written. For that quarter the report relies on the results release and the 8-K exhibit, and says so wherever it matters.

1. What the Company Does

Steel Dynamics melts scrap metal and turns it into new steel, and now aluminum too. It buys or collects old cars, factory offcuts, demolition steel and used cans. It melts them in electric furnaces and rolls the result into sheet, beams, rail and bar. Construction firms, carmakers, energy companies, equipment makers and steel distributors buy those products. It also owns the next steps in the chain. It runs one of North America's large scrap-collection networks (OmniSource), which feeds its own furnaces. It owns a fabrication business (New Millennium) that turns its steel into the joists and roof deck holding up warehouses and data centers. And since 2025 it owns a 650,000-tonne aluminum rolling mill in Mississippi that makes beverage-can sheet, automotive sheet and industrial sheet from recycled metal.

Founding and how it became what it is. Per the company's 2026 proxy statement, Mark D. Millett co-founded the company in 1993 and has been a director since then. He has been CEO since January 2012 and board chair since May 2021, and earlier held the roles of President and COO, EVP Metals Recycling and EVP Flat Roll Operations. Press accounts of the founding name Keith Busse and Richard Teets alongside Millett as co-founders. The 10-K dates construction of the first mill, in Butler, Indiana, to 1994, and the company listed in 1996. The idea from day one was the minimill: an electric arc furnace (EAF) that melts scrap. It needs a fraction of the capital of a blast furnace and can be idled or ramped with demand. Butler was a flat-rolled sheet mill, at the time an unusual product for a scrap-based mill.

Three later decisions shaped the company that exists today:

• Owning the scrap. Buying OmniSource in 2007 turned the company from a scrap buyer into a scrap supplier. About 65% of its ferrous scrap shipments in 2025 went to its own mills (FY2025 10-K).
• Going south and getting bigger. The Columbus, Mississippi flat-roll mill was acquired and expanded. The Sinton, Texas mill was built from scratch and started up in 2022. Sinton was built to serve Texas, the energy belt and Mexico without the freight cost of shipping from Indiana.
• Entering aluminum. In 2021 management decided to build a recycled aluminum flat-rolled mill next to its Columbus steel mill. The plan was to bring the scrap-and-EAF playbook into a market dominated by a few incumbents and short of domestic supply. First commercial coils shipped on 16 June 2025. That ramp is the central operating story of this report (Section 3).

In December 2025 the company also bought the remaining 55% of New Process Steel, a Houston-based processor and distributor. The company describes New Process as its single largest flat-roll steel customer, so that purchase moved Steel Dynamics a step closer to the end customer.

The problem it solves. Industrial buyers need steel, and increasingly aluminum, that arrives fast, meets specification and costs less than imported or blast-furnace metal. A scrap-fed EAF network spread across the Midwest, South and Texas lets the company quote short lead times, carry less inventory for its customers and offer metal with lower embedded carbon. Lower embedded carbon matters more and more to automotive, data center and consumer-packaging buyers.

What makes it hard to copy. A greenfield flat-rolled mill takes years and billions of dollars. Securing enough prime scrap to feed it is a second, separate problem. Qualifying products with automakers and can makers is a third. Steel Dynamics has solved all three at several sites at once. Section 3 covers the detail.

A concrete example. Take a data center developer in the Southeast. New Millennium designs and fabricates the roof joists and deck for the building shell from steel made at Columbus. OmniSource collects the developer's demolition scrap and a nearby auto plant's offcuts and sends them back into the same furnace. The company's processing and distribution arms (The Techs, Heartland, United Steel Supply and now New Process) coat, cut and deliver sheet to the contractors making HVAC units and electrical enclosures for the building. A single metals relationship covers material that would otherwise come from four or five suppliers.

2. Business Segments

Steel Dynamics reports four operating segments plus an "other" bucket. In FY2025 the mix of consolidated net sales was: Steel Operations 72%, Metals Recycling 11%, Steel Fabrication 8%, Aluminum Operations 2%, with the remainder in other businesses (FY2025 10-K). The mix is now shifting quickly as aluminum ramps. In the first nine months of 2026 the shares were roughly 67% steel, 10% recycling, 7% fabrication, 7% aluminum and 9% other (Q3 2026 release).

2.1 Steel Operations (about 72% of FY2025 sales)

What it does. The steel segment holds the EAF mills and their downstream processing:

• Three flat-roll mills: Butler, Indiana; Columbus, Mississippi; and the Southwest-Sinton Flat Roll Division in Texas.
• Four long-product mills: Structural and Rail Division in Columbia City, Indiana; Engineered Bar Products; Roanoke Bar; and Steel of West Virginia.
• Processing and finishing businesses: The Techs, Heartland, United Steel Supply and Vulcan Threaded Products.

Sheet (hot-rolled, cold-rolled and coated) was 64% of steel-segment sales in 2025, on 10.0 million tons of sheet production (FY2025 10-K). The rest is structural beams, railroad rail, special-bar-quality (SBQ) and merchant bar, and rebar. Customers range from service centers and pipe and tube makers to automotive suppliers, HVAC and appliance makers, railroads, equipment builders and construction contractors.

Core capability. The segment makes high value-added products from scrap: galvanized, Galvalume and painted sheet, rail, and engineered bar. That ground has traditionally belonged to integrated mills. Coating and painting capacity sits next to the hot mills, so the company sells a higher share of finished product than a plain hot-band producer does.

Role in the group. This is the earnings engine and the cash source that paid for aluminum. Management treats it as the base the other segments are built around.

Main rivals: Nucor, Cleveland-Cliffs, U.S. Steel (Nippon Steel), BlueScope's North Star, AM/NS Calvert and imports.

2.2 Metals Recycling (about 11%)

What it does. Under the OmniSource name, the segment collects, sorts and processes ferrous and nonferrous scrap across North America, including Mexico. In 2025 it shipped 6.16 million gross tons of ferrous scrap and 916.5 million pounds of nonferrous metal (FY2025 10-K). Its outside customers are other steel mills, foundries and nonferrous consumers. About 14% of the segment's 2025 sales were exports.

Core capability. It has dense collection and sorting next to its own mills. Increasingly it separates aluminum scrap by alloy, which is the input the new aluminum mill needs.

Why separate. It runs on different economics (volume times processing spread) and has outside customers. Strategically, though, it is a supply-security function: management repeatedly calls the Mexican yards a raw-material advantage for Columbus and Sinton (Q4 2025 call).

Main rivals: a fragmented field including SA Recycling, Radius Recycling (now part of Toyota Tsusho), CMC's recycling arm and Nucor's David J. Joseph.

2.3 Steel Fabrication (about 8%)

What it does. New Millennium Building Systems designs and makes steel joists, joist girders and steel deck at seven plants nationwide. It shipped 561,000 tons in 2025, and the company estimates it holds about one-third of domestic joist and deck bookings (FY2025 10-K). End markets are non-residential buildings: warehouses, distribution centers, manufacturing plants, data centers, schools and healthcare.

Core capability. It has engineering and detailing capacity, a national plant footprint and supply of sheet and bar from its own mills, so it can quote large projects quickly and reliably.

Why separate and how it fits. Fabrication earns from project pricing set months ahead, while its steel costs float. That makes it a partial hedge against the steel cycle. It is also a captive outlet for mill tonnage.

Main rivals: Nucor's Vulcraft/Verco (the largest), Canam and regional fabricators.

2.4 Aluminum Operations (about 2% of FY2025 sales, about 7% of 9M 2026)

What it does. Aluminum Dynamics runs the Columbus, Mississippi recycled aluminum flat-rolled mill (650,000 metric tonnes a year). It also runs a recycled slab center in San Luis Potosí, Mexico, and a deox-rod facility, and is building a second slab center in Mississippi (Section 3). At full production the targeted mix is roughly 45% beverage-can sheet, 35% automotive sheet and 20% industrial and common alloy (FY2025 10-K).

Core capability. The scrap-sorting and melt know-how from steel is being carried over to aluminum. The target is high recycled content: 85%+ for can sheet and about 65% for automotive and industrial (FY2025 10-K).

Why separate. It is a different metal with different customers, qualification regimes and pricing structure.

How it fits. This is the growth bet. Management has kept a through-cycle earnings target for the mill unchanged since construction and on the Q3 2025 call called the mill an option on a "larger asset" later (Section 7).

Main rivals: Novelis, Arconic, Constellium, Kaiser Aluminum and Tri-Arrows, plus imports.

2.5 Other

This bucket holds smaller businesses, including New Process Steel's processing operations after December 2025. It is not a separate strategic pillar.

Summary comparison

Geographic mix and export history

Steel Dynamics is overwhelmingly a domestic US business. Steel-segment exports were about 4% of segment sales in 2025, down from 6% in 2024 and 8% in 2023. Recycling exports were about 14%, down from 17% and 19% (FY2025 10-K). The falling export share reflects a domestic market tight enough to absorb everything the company makes.

Mexico is the one meaningful foreign footprint:

• OmniSource yards
• New Process Steel's two Mexican plants
• the San Luis Potosí aluminum slab center

Sinton was sited partly to reach northern Mexican manufacturing by rail. Mexico mostly serves as a scrap source and a near-shore customer base, not a separate market.

3. Products and Operations

3.1 How an EAF flat-roll mill works here

Each flat-roll division follows the same chain:

1. Charge. Shredded and prime scrap, blended with pig iron or other iron units, is loaded into the furnace. Iron units other than scrap were about 13% of metallics in 2025. Butler's Iron Dynamics unit produces liquid pig iron on site.
2. Melt. An electric arc furnace melts the charge. Electricity is about 4% of steel cost of goods sold (FY2025 10-K).
3. Refine and cast. Ladle metallurgy adjusts the chemistry, then a thin-slab caster casts the steel.
4. Hot roll. A tunnel furnace and hot strip mill roll the slab to hot band within minutes.
5. Finish. Hot band is pickled, cold-rolled, annealed, galvanized, Galvalume-coated or painted on site or at the downstream processing sites.

Thin-slab casting is the economic secret. Casting near final thickness cuts reheating energy and rolling passes, which is what lets a scrap mill compete with blast-furnace sheet. In total the company has about 16 million tons of steelmaking and coating capacity (FY2025 10-K) and shipped a record 13.7 million tons in 2025. In Q2 2026 the steel mills ran at about 90% utilization against a domestic industry average of about 81% (Q2 2026 call).

Sinton (Texas, started up 2022) was designed for wider and heavier-gauge coil than the earlier mills, aimed at energy, pipe and tube and southern manufacturing. Its ramp was slower than planned. The company now pairs it with a target through-cycle earnings contribution that it has not yet reached (Section 9).

Recent plant events:

• Transformer failure, January 2026. Barry Schneider, President and COO, said "damage was limited to the transformer itself, and operations resumed... within twelve hours" (Q4 2025 call).
• Four coating lines. Four new value-add coating lines added in 2024 ran at about 60% of capability through 2025 because of quality issues, which management said were fixed by early 2026 (Q4 2025 call).

3.2 Long products

• Structural and Rail Division (Columbia City, Indiana): wide-flange beams and railroad rail. The company is one of very few domestic rail producers, and Schneider has repeatedly called rail and structural demand "extremely strong" (Q3 2026 release).
• Engineered Bar Products: SBQ bar for automotive and heavy equipment, where metallurgical cleanliness and certification matter.
• Roanoke Bar and Steel of West Virginia: merchant bar, rebar and specialty sections for trailers, mining and industrial use.

3.3 Downstream processing and distribution

The Techs and Heartland coat and finish sheet, with about 2.0 million tons of processing capacity (FY2025 10-K). United Steel Supply distributes coated products. Vulcan makes threaded rod. New Process Steel, fully owned since December 2025, runs six manufacturing sites: four in the US, including sites at the Butler and Columbus divisions, and two in Mexico. It does blanking, slitting, laser cutting and supply-chain management for OEMs. A fatal accident at New Process in Q2 2026 led to a renewed safety review (Q2 2026 call).

3.4 Steel fabrication

New Millennium's seven plants take sheet and bar, much of it from the company's own mills, and cut, bend and weld them into custom joists and deck. Each job is engineered to the building. The constraint is labor and engineering throughput, not steel supply. Fabrication typically holds 10-12 weeks of inventory, so rising input costs squeeze margins until higher-priced backlog ships (Q2 2026 call).

3.5 Metals recycling

OmniSource yards buy obsolete scrap (cars, appliances, demolition) and industrial scrap (factory offcuts). They shred, shear and separate it, increasingly with sensor-based sorting, and ship it to mills. Nonferrous recovery (aluminum, copper, stainless) has become strategic because it feeds the aluminum mill.

3.6 The aluminum mill: the operation that matters most

Flow. Sorted aluminum scrap, topped up with prime metal, moves through four stages:

1. Melt and cast. The scrap is melted and cast into slab at the Columbus cast house and at satellite slab centers.
2. Hot roll. Slab is reheated and hot-rolled into "hot band" on a hot strip mill.
3. Cold roll. Three cold mills take the hot band to final gauge.
4. Finish. Each product has its own finishing route:
   • Can body and end stock is coated, leveled and slit.
   • Automotive sheet needs Continuous Annealing and Solution Heat (CASH) lines, which heat-treat and pretreat alloys for exposed and structural car panels.
   • Industrial and common-alloy sheet goes to building products, trailers and similar markets.

Satellite slab centers. Each satellite center is designed for 150,000 tonnes a year of recycled slab (FY2025 10-K). They let the company melt scrap close to where scrap is generated, which raises recycled content and lowers metal cost.

• San Luis Potosí, Mexico: cast its first ingot in March 2025.
• Arizona (cancelled): the second center was planned for Arizona and cancelled after disagreements with state officials. The company took an impairment of $16 million.
• Lowndes County, Mississippi (replacement): the center was moved to the Golden Triangle Industrial Park, next to Columbus, with start-up planned for the first half of 2027 (Q2 2026 call and Steel Market Update, July 2026).

The ramp so far.

In Q3 2026 yield losses rose at higher rolling speeds. The release traced this to "the hot strip mill rolling coolant composition, which has been successfully changed". The second CASH line is due to start in early Q1 2027 (Q3 2026 release).

Qualifications. Customer certifications came earlier than planned: for industrial sheet, for can sheet, and for automotive hot band (Q4 2025 call). Finished automotive product from the first CASH line was still in qualification at Q3 2026 and is expected to ship before year-end.

3.7 Biocarbon

SDI Biocarbon Solutions in Columbus, Mississippi is 75% owned by the company and 25% by Aymium. It pyrolyses sustainably sourced biomass into biocarbon to replace coal-based carbon in the EAF. Expected capacity is about 228,000 tonnes a year. First product shipped in September 2025 and was used at the Columbus steel mill (Q3 2025 call). It supports the company's lower-embodied-carbon BIOEDGE and EDGE steel product lines.

4. Customers

Who buys:

• Steel processors and service centers. Historically the largest channel. New Process Steel was the single largest flat-roll customer until the company bought it in December 2025.
• OEMs: automotive suppliers, HVAC and appliance makers, agricultural and heavy equipment builders, energy, pipe and tube producers, railroads.
• Non-residential contractors, mainly through New Millennium.
• National accounts for fabrication: large retailers, distribution center owners and data center builders (FY2025 10-K).
• Aluminum: can makers, automotive OEMs and their stampers, and industrial sheet users.

The company does not name individual large customers in its filings.

How each buyer type decides:

• Service centers and distributors. A purchasing manager buys monthly against inventory turns. The criteria are price, lead time, consistency of gauge and coating, and whether the mill will reliably allocate tonnage when the market tightens. The cycle is days to weeks. The company's advantage is proximity: three flat mills spread from Indiana to Texas put most of the Midwest, Southeast and Gulf within short freight distance. Owning New Process turns part of this channel into an internal customer.
• Automotive. Purchasing teams and engineering jointly approve each grade and part through formal qualification (PPAP-style approval for each grade and part). That takes months to more than a year, and pricing is usually set on longer contracts. Once a part is approved on a mill's steel or aluminum, switching means requalifying, which buyers avoid mid-model-cycle. That is real switching cost, and it is why the CASH-line qualification in Section 3.6 matters so much for aluminum.
• Can makers. The plant procurement and quality teams qualify sheet for each can line, and contracts usually run a year or more. Price is a conversion fee plus pass-through of the metal cost. Can makers want a second domestic source, especially after supply disruptions at an incumbent's plant in 2025 and the 50% tariffs on Canadian aluminum (Section 6). That desire for a second source is the opening Steel Dynamics walked through.
• Non-residential construction. The fabricator or general contractor awards joist and deck packages by bid, months before steel is needed. The criteria are engineering support, schedule certainty and price. New Millennium's one-third share and national footprint make it the obvious second bidder against Nucor's Vulcraft, and often the first.

Why they choose Steel Dynamics:

1. Short lead times from geographically spread mills.
2. Broad coated and painted product range from a single supplier.
3. Lower embodied carbon (scrap-based EAF, plus biocarbon) for buyers with emissions targets.
4. Reputation for allocation discipline in tight markets.

Switching costs. Switching costs are low for commodity sheet and bar, where a service center can move tonnage next month. They are moderate to high for automotive steel and aluminum grades and for can sheet once a line is qualified.

Concentration. No single external customer is disclosed as material. The largest single relationship, New Process, has been absorbed.

Contract structure. Steel is sold mainly on negotiated spot contracts priced monthly (FY2025 10-K), with longer-term pricing for some automotive volumes. Revenue is therefore highly price-sensitive within a few months. Fabrication is backlog-based: the backlog at Q3 2026 ran into Q2 2027 (Q3 2026 release), which gives visibility of two to three quarters. Aluminum, on management's design, will carry more contract business: conversion-fee pricing with metal pass-through. That is the basis for its claim that aluminum earnings will be steadier than steel.

5. Competitive Landscape and Moat

Industry structure. US flat-rolled steel is now an oligopoly of four large domestic groups:

• Nucor: EAF.
• Steel Dynamics: EAF.
• Cleveland-Cliffs: mostly blast furnace, with a heavy automotive mix.
• U.S. Steel: owned by Nippon Steel since June 2025, mixing integrated mills with the Big River EAF complex.

Smaller players sit around them: BlueScope's North Star mill in Ohio, AM/NS Calvert (the ArcelorMittal and Nippon Steel joint venture) in Alabama, and imports. The EAF producers have taken share from integrated mills for three decades because their costs are lower and more variable. Long products (beams, rail, bar, rebar) are split among Nucor, Steel Dynamics, CMC, Gerdau and a few others. Aluminum flat-rolled is more concentrated still. Novelis and Arconic dominate can and automotive sheet, with Constellium and Kaiser in specialty niches, and imports fill a structural shortfall.

How Steel Dynamics wins and loses against each:

• Nucor. The closest copy: same technology, similar scrap integration, a rival joist business. Steel Dynamics wins on concentration (fewer, larger, newer flat mills) and on painted and coated mix. Nucor wins on breadth, plate and balance sheet. Both now add flat capacity into the same market (Section 6).
• Cliffs and U.S. Steel. Their integrated mills carry fixed costs, coke and iron-ore exposure, and pension legacies. Steel Dynamics underprices them on commodity and coated sheet. They keep the edge in exposed automotive grades, where Cliffs has decades of qualifications.
• North Star. Matches Steel Dynamics on hot-band cost in the Midwest but lacks coating breadth in the US. The company's repeated attempts to buy it (Section 8.6) say a lot about how management rates that asset.
• Imports. Effectively shut out of most products by the 50% Section 232 tariff (Section 6). Their return depends on policy, not on competition.
• Aluminum incumbents. Novelis and Arconic have decades of automotive qualifications and customer relationships. Steel Dynamics enters with a newer mill, scrap integration and a can-maker base eager for a second domestic source. It is behind on automotive and still climbing the learning curve on yields (Section 3.6).

Barriers to entry. They are high. A flat-rolled EAF mill costs several billion dollars and takes about three years to build, and then two to three more to qualify products and reach full rates. Prime scrap and pig iron supply is finite and contested. Qualifying products with automakers and can makers adds a time barrier capital cannot buy. Even so, barriers have not stopped incumbents from expanding: Nucor West Virginia, Big River 2 and Novelis Bay Minette are all incumbent capacity additions.

Share distribution. Steel Dynamics is the second-largest US EAF producer behind Nucor. The concentration reflects decades of consolidation and the closure of integrated capacity. The company's own estimate of about one-third of domestic joist and deck bookings makes New Millennium one of two national players.

Structural shifts:

1. EAF share keeps rising as integrated mills close or convert.
2. A new wave of EAF flat capacity, the Steel Dynamics expansions included, raises the risk of oversupply when demand softens (Section 8.2).
3. Japanese ownership of U.S. Steel puts a deep-pocketed technology partner behind a rival.
4. In aluminum, two new domestic mills, Steel Dynamics and Novelis Bay Minette, will narrow the import gap from 2027.

Moat: Narrow - scale / cost advantage, switching costs (in qualified automotive and can grades)

The company makes most of its steel from scrap it partly collects itself, in newer thin-slab mills spread across the main US demand regions, with a high coated and painted mix. That gives it a structural cost and freight position integrated rivals cannot match (Sections 3 and 4). Qualified automotive and can grades add moderate switching costs. The advantage is still narrow. Nucor runs the same model at larger scale, so the cost edge is shared rather than unique. Most tonnage is sold spot, where customers switch freely. And the aluminum franchise has not yet proved itself through a downturn. The single biggest threat to the moat is the 2027 wave of new domestic EAF flat-rolled capacity, which will compete on the same cost basis rather than against higher-cost integrated mills.

6. Industry

What drives demand. US steel demand follows non-residential construction (warehouses, factories, data centers, healthcare), automotive production, energy (pipelines, solar, transmission, oil and gas tubulars), agricultural and heavy equipment, and infrastructure. The company named energy, non-residential construction, automotive and industrial as the sectors leading demand in Q3 2026. Its fabrication backlog was driven by "commercial construction, manufacturing, warehouses, data centers, and healthcare" (Q3 2026 release). Aluminum flat-rolled demand follows beverage packaging (a structural shift from plastic and glass to cans), vehicle lightweighting and building products. On the Q3 2025 call management put the structural North American aluminum-sheet deficit at more than 1 million tonnes a year before tariffs.

Size and trajectory. Imports supplied about 16% of the US steel market in the first nine months of 2026, down from about 19% a year earlier. US rolled-steel imports fell about 27% year on year in January to May 2026 (IndexBox, AISI data). Domestic producers are filling that gap. Mill utilization of about 81% for the industry (Q2 2026 call) leaves room to produce more without new capacity, and new capacity is arriving anyway:

• Big River 2: U.S. Steel's Arkansas complex is now online, doubling site capacity to about 6 Mt.
• Nucor West Virginia: a 3 Mt sheet mill commissioning in late 2026, with commercial shipments from early 2027.

In aluminum, Novelis made the first coil on its Bay Minette, Alabama cold mill in October 2026 and is targeting commercial shipments in 2027. The plant has about 600,000 tonnes of recycled capacity.

Where Steel Dynamics sits in the supply chain. It sits in the middle, converting scrap into finished and semi-finished metal, and integrates backward into scrap and forward into processing, distribution and fabrication. Globally, the US is a high-price island: domestic prices sit well above world levels because the market is protected.

Trade and regulation. Trade policy is the single largest industry-level variable:

• Section 232 tariffs on steel and aluminum imports were restored and broadened in 2025 and raised to 50% in mid-2025. Executive orders in April 2026 extended coverage across derivative products, which management called "very helpful because they encompass the entire supply chain" (Q1 2026 call).
• Section 301 investigations of countries with excess steel capacity were under way in 2026. The domestic industry is lobbying for any remedies to be "additive to the existing 232 steel tariffs" (Q2 2026 call).
• Brazilian pig iron. A Section 301 action against Brazil excluded pig iron imports, which affects metallics costs for all EAF producers.
• Aluminum. A 50% tariff, including on Canadian metal, has hurt cross-border suppliers such as Novelis's Kingston, Ontario plant and pushed can makers toward domestic supply.

Import substitution. For aluminum sheet, the new domestic mills are explicitly import-substitution projects. For steel, low imports are a policy outcome rather than a cost outcome. They last only as long as the tariffs do.

Cyclicality. Steel is deeply cyclical. Prices can swing by half within a year as service-center inventories build and drain. Spot pricing passes those swings straight into mill earnings. Scrap prices move with steel prices, so EAF margins (the "metal spread") swing less than prices do, but still a lot. Construction-linked fabrication lags the cycle by two to three quarters. Can sheet is the least cyclical metal product, because beverage demand is steady and conversion-fee contracts pass metal costs through.

Tailwinds. Reshoring of manufacturing, data center construction, grid and energy investment, infrastructure spending, low customer inventories, tariff protection, and buyer preference for lower-carbon metal.

Headwinds.

• New domestic flat-rolled and aluminum capacity arriving in 2027.
• Competition among EAF producers for prime scrap.
• Interest-rate sensitivity of construction and autos.
• Dependence on tariff policy that a future administration or court could change.

7. Growth Triggers

Ranked by how much each could change the business.

• Aluminum mill to full rate in 2027. In January 2026 management raised its end-2026 exit-rate target from 75% to "approaching 90%" of the 650,000-tonne capability, citing early certifications and equipment redundancy (Q4 2025 concall, Jan 2026). It repeated the target in July: the mill would "exit 2026 at a monthly production rate of at least 90% capacity, allowing for full volume capability in 2027" (Q2 2026 concall, 21 Jul 2026). In October all three cold mills were running, giving "the ability to achieve our annual volume expectations" (Q3 2026 release, 19 Oct 2026). Repeated across four results events.
• Automotive aluminum via the CASH lines. The first CASH line is operating and expected to ship finished automotive product before the end of 2026. The second is to start in early Q1 2027 (Q3 2026 release, 19 Oct 2026; previewed in the 17 Sep 2026 guidance release). Management has guided to a richer, higher-margin mix as automotive volumes arrive, "reaching optimization sometime in 2027" (Q3 2025 concall, 21 Oct 2025).
• Aluminum turning profitable and staying there. Management expects aluminum to move "from EBITDA positive to earnings positive in the second half of 2026" (Q2 2026 concall, 21 Jul 2026). It expects volumes and profitability to improve in Q4 2026 and through 2027 "as startup costs subside, utilization and yields improve, and scrap content increases" (Q3 2026 release). The through-cycle earnings target for the mill and for the aluminum side of recycling has been reaffirmed on every call since Q2 2025.
• Second recycled slab center, Mississippi, H1 2027. The relocated slab center at the Golden Triangle Industrial Park is to start operating in the first half of 2027. It raises recycled scrap content and lowers metal cost, and brings the second cast house to "full utilization in the first half of 2027" (Q2 2026 concall, 21 Jul 2026).
• Steel fabrication backlog repricing. Backlog was 45% higher year on year in July and 50% higher in October, extending into Q2 2027 (Q2 2026 concall; Q3 2026 release). CFO Theresa Wagler said the higher pricing in that backlog will show up "over the next six to nine months" (Q2 2026 concall). Strong order entry was first flagged in January and April (Q4 2025 and Q1 2026 concalls), so this has been repeated four times.
• Sinton and the four coating lines approaching design earnings. Management keeps a through-cycle earnings target for Sinton and a separate one for the four value-add coating lines. It says the coating lines' quality problems were fixed by early 2026, after running at about 60% of capability in 2025 (Q4 2025 concall, Jan 2026; Sinton target repeated Q3 2025 concall, 21 Oct 2025).
• Capital spending falls back after the build-out, freeing cash. Management expects 2027 capital spending to be at most roughly the 2026 level, with sustaining spending running at about half of it (Q2 2026 concall, 21 Jul 2026). Wagler tied any dividend step-up to "structural cash flow changes" (Q4 2025 concall, Jan 2026), and buybacks "could trend near" the Q2 2026 pace (Q2 2026 concall).
• A larger aluminum asset (option, not a plan). "There is definitely growth opportunity there. Both in downstream... and... there's still clear room for a larger asset, the mill asset" (Mark Millett, Q3 2025 concall, 21 Oct 2025). He framed it with "we want to make sure we walk before we run". On the Q4 2025 call Wagler said no specific material projects had been named beyond 2026.

8. Key Risks

8.1 The aluminum ramp keeps slipping, and arrives into tougher competition. The mill has had a different quality problem almost every half-year. Certification work in Q3 2025 cost more than expected. Product staining hit in January 2026. Rolling-coolant yield losses followed in Q3 2026. Each was fixed, but each pushed volume later. Millett was candid about the staining:

"It was a quality issue. It was a stain on the product. We should have caught it." - Mark Millett, Q1 2026 call, 21 April 2026

The mechanism of harm is time:

• Every quarter at partial utilization spreads a large fixed-cost base over fewer tonnes.
• Automotive qualification on the CASH lines can take longer than planned.
• Novelis's Bay Minette mill (Section 6) begins commercial shipments in 2027, the same year Steel Dynamics expects full volume, so can-sheet conversion fees could narrow just as the mill arrives.

Calibration: high probability of moderate further delay. Failure of the asset is a low-probability outcome. Section 9 shows every aluminum timeline so far has run late.

8.2 A domestic flat-rolled supply wave meets a softening market. About 9 Mt of new or recent EAF flat capacity is landing in the US market: Big River 2, Nucor West Virginia and the company's own Sinton and coating lines. While imports are low and inventories lean, the market absorbs it. If construction or autos slow in 2027, spot pricing would spread the price cut across almost all of the company's tonnage within a quarter or two. Calibration: moderate probability, potentially large earnings swing. This is the normal steel cycle, made sharper by the capacity additions.

8.3 Trade protection is reversed or diluted. Today's market tightness depends on the 50% Section 232 tariff, its extension to derivative products, and possible Section 301 add-ons, all of which management lobbies to preserve (Q2 2026 call). Tariff-rate deals with allied countries, court challenges or a change of administration could let imports back in. Calibration: low-to-moderate probability within two to three years, high impact, because imports could quickly retake several points of share from today's 16%.

8.4 Metallics cost squeeze. Every new EAF chases the same prime scrap and pig iron. Owning OmniSource protects volume, not price. If scrap rises faster than steel prices, the metal spread compresses across the company's EAF base. The new aluminum mill adds a second scrap market in which the company competes with its own customers' recyclers. Calibration: moderate probability, moderate drag. It is partly hedged, because recycling earns more when scrap prices rise.

8.5 Concentration of growth assets at two Gulf-region sites. Sinton and Columbus hold the newest flat-rolled steel, the entire aluminum business, the biocarbon plant and the new slab center. Both sit in hurricane-exposed regions with single-point equipment risks, such as Sinton's January 2026 transformer failure (Section 3.1). A prolonged outage at either site would hit the growth story disproportionately. Calibration: low probability, high impact.

8.6 Large-deal appetite and leadership concentration.

The BlueScope bids. In January 2026 the company and Australia's SGH jointly proposed to buy BlueScope. SGH would keep the Australian and Asian businesses and sell BlueScope's North American operations, including the North Star mill and building and coated products, to Steel Dynamics. BlueScope's board rejected the first proposal on 7 January 2026. A revised "best and final" A$32.35-per-share proposal in February was also rejected. BlueScope had by then rejected five approaches, the first of them made in late 2024 (BlueScope ASX release; Bloomberg). Millett said in April that there had been "no constructive engagement by the company since" (Q1 2026 call). The bid is dormant, not withdrawn. The risk runs both ways. A future deal at a peak-cycle price would load the balance sheet just as aluminum and Sinton are meant to start paying back. Persistent interest also shows management is willing to pay up for North American steel assets.

Leadership. Strategy and execution are concentrated in co-founder Mark Millett, Chairman and CEO and 66 per the 2026 proxy, and in a small top team. The board reports regular succession discussions, but no successor has been publicly named (2026 proxy). Calibration: low-to-moderate probability, moderate-to-high impact.

9. Walk the Talk

The six results events used:

1. Q2 2025 call, 22 July 2025
2. Q3 2025 call, 21 October 2025
3. Q4 2025 call, late January 2026
4. Q1 2026 call, 21 April 2026
5. Q2 2026 call, 21 July 2026
6. Q3 2026 results release, 19 October 2026, with the 17 September 2026 guidance release

The Q3 2026 call was held on 20 October 2026, but no transcript was available when this was written. Q3 judgements rest on the company's written releases.

The aluminum timeline is where the misses sit. Every aluminum volume and loss target set in 2025 and early 2026 has slipped by one or two quarters. The explanations changed each time:

• Q3 2025: certification costs.
• Q1 2026: a staining problem that, in Wagler's words, "was not an equipment issue... it was a process issue".
• Q3 2026: rolling-coolant chemistry. The Q3 2026 release put it this way:

"Volumes were less than expected over a number of weeks, as yield losses increased at higher operating speeds due to quality impacts. The cause originated from the hot strip mill rolling coolant composition, which has been successfully changed, and no ongoing additional costs are expected." - Q3 2026 results release, 19 October 2026

What management did not do is lower the end-state targets. In January 2026, with December shipments at about a fifth of capability, it raised the 2026 exit target from 75% to about 90%. Nine months later the mill is still well short of that rate with one quarter to go. When analysts in July asked whether the long-run aluminum earnings target should go up given strong market prices, Wagler declined:

"We want to make sure that we actually have everything dialed in before we start to talk about numbers that are higher." - Theresa Wagler, CFO, Q2 2026 call, 21 July 2026

That restraint on the end-state number sits oddly next to repeatedly optimistic near-term ramp dates.

On steel and fabrication, the record is better. Calls about demand, backlog and pricing have been close to what happened. December 2025 was flagged as setting up "2026 very well", and steel shipments then set records in Q1, Q2 and Q3 2026. The September 2026 earnings guidance missed by a small margin rather than a large one.

Assessment. On the core steel business and on capital returns, management does what it says. On new-asset start-ups it is consistently optimistic about timing, but not about the end result: completion targets have stood while ramp dates slipped. The aluminum ramp dates are best read as best-case schedules.

10. Ownership, Governance and Shareholder Friendliness

Ownership and control. Steel Dynamics has a single class of common stock and no controlling holder. About 144.7 million shares were outstanding at the 16 March 2026 record date (2026 proxy).

Insiders, including co-founder Millett, own low single digits. For a minority shareholder the structure is plain: one share, one vote, a wide free float, and control exercised through the board and widely held institutional votes rather than by a family or parent.

Governance flags:

• Board. Seven director nominees, six of them independent. The CEO also chairs the board. Kenneth W. Cornew has been Lead Independent Director since 2023 (2026 proxy).
• Related-party transactions are disclosed and small. They are relatives of executives employed below executive level, each paid under $475,000:
  • a brother-in-law of Mark D. Millett, who is a sales manager at Butler;
  • a brother of Christopher A. Graham, SVP Flat Roll Steel, who works at Sinton;
  • a brother of retired SVP Glenn A. Pushis, who works in aluminum.
• Union Pacific dealings. Director Jennifer L. Hamann is CFO of Union Pacific. In 2025 Union Pacific paid the company about $44 million for rail and the company paid Union Pacific about $102 million for freight, described as "made on market terms" (2026 proxy).
• Pay. About 84% of the CEO's target pay is at risk. It is tied to a net-income-based bonus pool, return on equity, operating cash flow and relative total shareholder return against steel peers (2026 proxy).
• Hedging and pledging. Hedging is prohibited and pledging severely limited.
• Overall. No auditor change or qualified opinion was identified. Governance is unremarkable apart from the combined chair and CEO role.

Dividends.

The dividend has grown every year in the window (company dividend releases, February 2025 and 2026). Management describes it as a "base positive dividend profile" with variable buybacks on top (Q4 2025 call).

Buybacks and dilution.

Older programmes (outside MoatMap's window):

Sources: FY2023 and FY2024 results releases; Q4 2025 call; Q3 2026 release.

The last ~90 days. MoatMap's buyback window covers only the trailing ~90 days (since 22 July 2026). It shows one issuer filing, the 28 July 2026 quarterly report, which lists repurchase consideration rows but no share counts. Company figures show $276 million repurchased in Q3 2026, about 1% of the stock, $261 million of it early in the quarter (Q3 2026 release; 17 September 2026 guidance release).

Net effect on share count. Shares outstanding fell from 171.6 million at 21 February 2023 (FY2022 10-K cover) to 144.7 million at 16 March 2026 (2026 proxy), a net reduction of about 16% in three years, with no meaningful option dilution.

Verdict: Returns Capital. The company cut its share count by about a sixth in three years and raised its dividend every year, while also paying for the aluminum mill and Sinton.

11. Insider Activities

The spine for this section is MoatMap's nightly scrape of SEC Form 4 filings, current to 20 October 2026. The most recent two weeks were checked against EDGAR Form 4 listings. The only filing in that window is a 9 October 2026 Form 4 batch for directors, which matches the MoatMap rows, and no newer filings were found.

Buys. There were no open-market purchases by any director or officer in the 12-month window.

Sells. Three senior vice presidents made the only market sales: Anderson, Graham and Poinsatte, about 21,300 shares and roughly US$5.4 million in total. All fell between 12 May and 5 June 2026, as the stock rose from the mid-$230s to the high-$260s after the Q1 results. The pattern fits post-vesting diversification after the March award vesting and the opening of the post-earnings trading window. Neither MoatMap's data nor the filings reviewed disclose the reason or a 10b5-1 plan, so the reason is formally not disclosed. Each sale is small next to the share count, and each seller kept a holding. The CEO, CFO and President made no market sales. Their only reported activity is tax withholding on vesting and Millett's small no-consideration disposition.

Net assessment. Insiders were net sellers in the market, but modestly. Selling was confined to three non-board operating executives in a single post-earnings window, with no CEO, CFO or director selling and no buying anywhere. Nothing changed recently: there has been no unusual first-time buyer and no cluster around the Q3 2026 release. Read: neutral.

12. Scenarios

Bull case. The coolant fix holds and the Columbus aluminum mill runs near full rate through 2027. The first CASH line finishes automotive qualification on schedule, the second follows in Q1, and the Mississippi slab center lifts scrap content in the second half. With can makers wanting a second domestic source and tariffs keeping Canadian and overseas sheet out, the mill fills a contracted, conversion-fee book before Novelis's Bay Minette arrives in force. Aluminum then becomes the steady earner management promised. In steel, the fabrication backlog reprices through 2027. Data center, reshoring and energy construction absorb the new capacity from Nucor West Virginia and Big River 2. Imports stay low under a durable Section 232 regime. With heavy capital spending finished, the company keeps shrinking its share count and steps up the dividend. By 2028 it looks less like a cyclical steelmaker and more like a two-metal recycling and rolling platform with steadier earnings.

Base case. The aluminum mill reaches full volume during 2027 rather than exiting 2026 there. That fits the Section 9 pattern of stated end-points arriving a quarter or two late. Automotive sheet ramps more slowly than can and industrial sheet, so the richer mix shows up late in 2027. Steel stays strong into the first half of 2027 on lean inventories and the long fabrication backlog. Prices then soften as new domestic capacity lands, and the metal spread narrows without collapsing. Sinton and the coating lines edge toward, but do not reliably hold, their design earnings. Capital returns continue at roughly the 2026 pace. The BlueScope interest stays dormant. In two to three years the company is recognisably what management described, just a year later than the original slides.

Bear case. 8.2 and 8.3 compound. New Nucor and U.S. Steel sheet capacity floods a market that is slowing on rates and construction. At the same moment a trade deal or court ruling loosens Section 232, and imports climb back toward a fifth of the market. Monthly spot pricing passes the decline into earnings within two quarters, and 8.4 tightens the squeeze as scrap holds up better than sheet prices. Meanwhile 8.1 plays out: another quality problem keeps aluminum below full rate into late 2027, just as Bay Minette competes for the same can makers. The mill's fixed costs then weigh on group results through a steel downturn. If 8.6 adds a large acquisition near the cycle peak, the company enters the trough with more leverage, slower buybacks, and two unproven growth assets on the Gulf Coast exposed to 8.5.

Sources: Steel Dynamics Q3 2026 results release and 8-K exhibit; Q3 2026 guidance release; Q2 2026 call transcript; Q2 2026 call summary; Q1 2026 call transcript; Q4 2025 call transcripts (Motley Fool, Investing.com); Q3 2025 call transcript; Q2 2025 call transcript; FY2025 Form 10-K; 2026 proxy statement; FY2022 10-K; Vanguard 13G; BlackRock holding; 2026 dividend increase; 2025 dividend and buyback authorization; FY2024 results; FY2023 results; New Process Steel acquisition; BlueScope rejection (ASX); best-and-final proposal; revised bid rejected; slab center relocation; biocarbon; founding history; Big River 2; US rolled-steel imports; US steel market and import share; Novelis Bay Minette; Section 232 tariffs; peer market caps from CompaniesMarketCap, StockAnalysis (CLF), StockAnalysis (CMC), StockAnalysis (CSTM) and CompaniesMarketCap (BlueScope); Q3 2026 earnings recap; Oct 2026 Form 4; MoatMap insider and buyback database (SEC Form 4 / 10-Q), current to 2026-10-20.