Steel Sector Outlook — September 2026

Read the report

SlidesScroll

1 of 22

←→Full screen

MATERIALS › MATERIALS › STEEL

Steel: Valuation Gaps Across This Set Are Not Tracking Current Operating Performance

A read of eleven listed steelmakers, what forward EBITDA separates inside the set, and what buyers agreed to pay in the recorded transactions.

September 2026 · Prepared by NeuraCap AI · Confidential

Market data as of 2026-09-28 · primary valuation basis EV / EBITDA (CY2027E)

Steel Coverage | September 2026 | Confidential | Not investment advice

Sources & methodology

1

← Previous

1 / 22 · Steel: Valuation Gaps Across This Set Are Not Tracking Current Operating Performance

Next →

What each page shows

The analyst’s walkthrough of the deck, page by page.

1. 01
   MATERIALS › MATERIALS › STEEL

   Steel: Valuation Gaps Across This Set Are Not Tracking Current Operating Performance

   Cover slide introducing the steel sector valuation report as of September 2026.

   We're opening with the headline finding: valuation gaps across this steel set are not tracking current operating performance. Over the next few pages we'll show where the premium sits and what it's associated with.

   Everything on this page

   MATERIALS › MATERIALS › STEEL Steel: Valuation Gaps Across This Set Are Not Tracking Current Operating Performance A read of eleven listed steelmakers, what forward EBITDA separates inside the set, and what buyers agreed to pay in the recorded transactions. September 2026 · Prepared by NeuraCap AI · Confidential Market data as of 2026-09-28 · primary valuation basis EV / EBITDA (CY2027E) Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 1

2. 02
   CONTENTS

   What This Report Covers

   Table of contents listing the five sections and appendix covered in the report.

   This report runs five sections plus an appendix — the bottom line, the landscape, valuation and situations, precedent transactions, and strategic implications. We've put the bottom line first, so if you only have time for one section, that's the one to read. Everything after it builds the evidence for that finding.

   Everything on this page

   CONTENTS What This Report Covers 01 The Bottom Line The Bottom Line: Price and Performance Sit Apart in Steel 02 The Landscape One Producer Group Holds the Whole Steel Screen 03 Valuation & Situations The Top and the Bottom of the CY2027E Range 04 Precedent Transactions What Buyers Agreed to Pay Across Nine Recorded Steel Transactions 05 Strategic Implications What the Gap Between the Two Ends Means for an Operator 06 Appendix Comparables Detail, Methodology, Sources and Assumptions Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 2

3. 03
   01 · THE BOTTOM LINE

   Steel Is Priced as One Field of Multi-Route Producers, and the Higher-Priced Half Is Not the Faster-Growing Half

   Summarizes the bottom line: pricing spreads inside one business model rather than between growth profiles.

   Every company in this set falls into one diversified multi-route producer classification, so the price differences we're about to show sit inside a single business model. On CY2027E EBITDA, the sector median runs at 5.7x, and the higher-priced half of the set is not the faster-growing half. That's the tension this report is built to explain, and it's why we start here before walking through the landscape and the deals.

   Everything on this page

   01 · THE BOTTOM LINE Steel Is Priced as One Field of Multi-Route Producers, and the Higher-Priced Half Is Not the Faster-Growing Half The full story on one page · figures on EV / EBITDA (CY2027E), market data as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Primary valuation basis: EV / EBITDA on CY2027E consensus (10 of 11 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Qualitative characterisations are NeuraCap views. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 3 1 The Price Gaps Here Open Inside One Business Model 11 of the 11 companies classify as diversified multi-route steel producers, 100% of the set, so the price gaps here sit inside one business model rather than between different ones. The middle of the range is 5.7x forward EBITDA on CY2027E, with 10 of the 11 companies carrying an estimate. 2 The Faster-Growing Half Carries the Lower Price Split at 4% revenue growth, the five faster-growing names sit at 5.3x and the five slower-growing names at 6.5x on CY2027E EBITDA. The gap runs against the usual direction, and a forward multiple already credits forecast growth, so the premium at the top is sitting with something other than the growth line. 3 The Price Gap Is Wider than the Performance Gap The three names at the top of the range sit at 7.4x on CY2027E EBITDA against 4.3x for the three at the bottom. The operating numbers underneath do not spread nearly as far, so the wider valuation gap is associated with views on how durable the earnings are judged to be rather than with current performance. 4 What Buyers Agreed to Pay Spans a Wide Band Across the nine recorded transactions, disclosed multiples run as high as 13.5x on an announced stainless long products carve-out. Cleveland-Cliffs Inc. agreed $9.2B for United States Steel Corporation, the largest disclosed value among the nine, with smaller cross-border carve-outs at the other end of the band. 5.7x Sector median EV/EBITDA CY2027E consensus · 10 rated of 11 companies 7.4x Premium end EV/EBITDA vs 4.3x at the discount end top quartile (n=3) against bottom quartile (n=3) on EV/EBITDA — the spread the report explains 21 Transactions with disclosed terms 58 recorded in this tier · 1 told as case studies, the full list in the appendix

4. 04
   SECTION 02

   02

   Divider introducing Section 02, the market landscape for the eleven-company steel set.

   This section maps where all eleven companies sit — one classification, one screen, and a wide price band inside it. We'll use it to reset before walking through the group-level view.

   Everything on this page

   SECTION 02 02 THE LANDSCAPE One Producer Group Holds the Whole Steel Screen Eleven companies, a single diversified multi-route classification, and a wide price band inside it. 02 of 06 Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 4

5. 05
   02 · MARKET MAP

   This Set Sits in One Group, and the Price Still Spreads Wide Inside It

   Shows all eleven approved companies grouped by business segment with group median EV/EBITDA.

   All eleven companies in this set land in one business-segment group, and the median EV/EBITDA on CY2027E inside that group still spreads wide. That tells us the segment label alone doesn't explain the price differences we're seeing — something else is driving the spread. We'll get into what that is on the next page.

   Everything on this page

   02 · MARKET MAP This Set Sits in One Group, and the Price Still Spreads Wide Inside It 11 approved companies grouped by business segment · median EV / EBITDA (CY2027E) per group · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Segment grouping follows the platform's classification; group medians on rated names only. "Why it matters" lines are NeuraCap views. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 5 DIVERSIFIED MULTI-ROUTE STEEL PRODUCERS 11 cos median 5.7x ArcelorMittal S.A. (MT) Nucor (NUE) Steel Dynamics (STLD) POSCO Holdings (PKX) Ternium S.A. (TX) Cleveland-Cliffs (CLF) Gerdau S.A. (GGB) Commercial Metals (CMC) Companhia (SID) Algoma Steel Group (ASTL) Metallus (MTUS) The set classifies into a single group of 11 producers, so what separates prices here is cost curve position, downstream mix and contract book rather than category.

6. 06
   02 · LANDSCAPE

   The Segment Label Does Little to Separate These Names; Operating Position Sits with the Price Gap

   Compares what each company does against where it prices, showing the segment label doesn't separate operating position.

   The segment tag groups every name the same way, but operating position — what each company actually does and how it's positioned — is where the price gap really sits. We're using this view to move the conversation from classification to substance. Full company-level detail is in the appendix if you want to trace any single name.

   Everything on this page

   02 · LANDSCAPE The Segment Label Does Little to Separate These Names; Operating Position Sits with the Price Gap Segment view of the approved universe · EV / EBITDA (CY2027E) medians on rated names · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. "What they do / why it matters" is a NeuraCap view. Full company-level detail in the appendix and companion tables. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 6 Segment n Share of universe Median EV/EBITDA Names to know What they do — and why it matters Diversified multi-route steel producers 11 100% 5.7x ArcelorMittal S.A. (MT) · Nucor Corporation (NUE) · +9 more One group, eleven operators. These producers run blast furnace–BOF and electric arc furnace routes across long products, coated and special bar quality mix. With 10 of the 11 companies carrying a CY2027E EBITDA estimate, this group is the pricing reference for any steel asset discussion, and the spread inside it is where the argument actually happens.

7. 07
   SECTION 03

   03

   Divider introducing Section 03, covering public market valuation across the CY2027E range.

   Ten of the eleven companies carry a CY2027E EBITDA estimate, and this section ranks those ten from top to bottom. We'll use it to show exactly where the premium and the discount sit.

   Everything on this page

   SECTION 03 03 VALUATION & SITUATIONS The Top and the Bottom of the CY2027E Range 10 of the 11 companies carry a CY2027E EBITDA estimate; here is how the 10 rated names rank. 03 of 06 Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 7

8. 08
   03 · PUBLIC MARKET VALUATION

   The Top of the Range Holds the Steadier Earners, the Bottom Holds the Faster Growers

   Ranks all ten rated companies by EV/EBITDA on CY2027E, sorted descending against the 5.7x sector median.

   Sorting the ten rated names from highest to lowest multiple, the top of the range holds the steadier earners and the bottom holds the faster growers — against a sector median of 5.7x. That ordering runs counter to the usual assumption that growth commands the premium. So the question worth asking is what the market is actually pricing at the top of this range.

   Everything on this page

   03 · PUBLIC MARKET VALUATION The Top of the Range Holds the Steadier Earners, the Bottom Holds the Faster Growers EV / EBITDA (CY2027E) · all 10 rated companies, sorted descending · sector median 5.7x · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Primary valuation basis: EV / EBITDA on CY2027E consensus (10 of 11 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). Tier zones are NeuraCap groupings cut at the rated set's quartiles; every multiple quoted on this page is a median on the same EV / EBITDA (CY2027E) basis. Panel commentary is a NeuraCap view. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 8 PREMIUM · median 7.4x CORE · median 5.7x DISCOUNT · median 4.3x Sector median 5.7x WHAT SEPARATES THE TWO ENDS The top holds 7.4x. The three names at the premium end sit at 7.4x on CY2027E EBITDA. Two of them post margins of 21% and 20% on revenue growth of 1% and 2%, so the price at the top of the range sits alongside steady profitability rather than volume expansion. The bottom holds 4.3x. The three names at the discount end sit at 4.3x on the same lens. Companhia Siderúrgica Nacional (SID) carries a 27% margin and still prices at the bottom of the range, which is why profitability on its own is not explaining the ranking on this page. Forward pricing already credits growth. The lens is CY2027E EBITDA, so a forecast of better shipments and better metal spread is already inside the number. A premium that survives that test points to earnings the market expects to hold through the cycle, and that is what the underwriting conversation turns on.

9. 09
   03 · VALUATION DRIVERS

   Profitability Separates the Two Ends: Names Above the 17% Margin Line Carry 4.5x Against 6.0x Below It

   Splits the rated names into growth and margin cohorts and compares median EV/EBITDA across each split.

   Splitting the set at a 17% EBITDA margin line, names above it carry 4.5x against 6.0x below it — margin cohorts move in the opposite direction from what we'd expect if margin alone commanded the premium. We read this as an association in the data, not a causal claim. It tells us profitability by itself isn't the whole story behind where these names price.

   Everything on this page

   03 · VALUATION DRIVERS Profitability Separates the Two Ends: Names Above the 17% Margin Line Carry 4.5x Against 6.0x Below It Median EV / EBITDA (CY2027E) by revenue-growth cohort and by EBITDA-margin cohort (each split at its covered median) · rated names with estimates · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Cohort medians on rated names with the required estimates (faster n=5; slower n=5; higher-margin n=5; lower-margin n=5). Driver readings are NeuraCap views on the supplied data — association, not causation. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 9 EV/EBITDA, median per cohort · growth split at 4% · EBITDA-margin split at 17% The Growth Split Runs Backwards Split at 4% revenue growth, the five faster-growing names sit at 5.3x and the five slower-growing names at 6.5x on CY2027E EBITDA. Both halves sit in the same producer group, so the gap is inside one business model rather than between two different ones. High Margin Shows up at Both Ends The margin line across the set runs at 17%. Companhia Siderúrgica Nacional (SID) at 27% and Ternium S.A. (TX) at 19% both price in the discount tier, while Nucor Corporation (NUE) at 20% prices in the premium tier, so margin on its own is not separating the field either. Earnings Dispersion Is the Live Variable Inside one producer group the visible differences are contract versus spot share of the order book, coated and special bar quality mix, and conversion cost per tonne. These are the items associated with EBITDA per tonne holding when metal spread moves, and they are what a buyer tests behind any multiple. Region Travels with the Price The three names at the bottom of the range are Latin American producers; the three at the top are North American. Freight economics, import competition and trade remedy regimes make those two positions different underwriting problems, on a set of 10 rated names rather than a census.

10. 10
    03 · SITUATION MAP

    Where the Money Sits: Price Against Growth for the 10 Rated Names

    Plots the ten rated names on price versus growth, cut at the sector median multiple and the covered median growth rate.

    Cutting the set on EV/EBITDA against the 5.7x sector median and on revenue growth against the 4% covered median gives us four situations rather than one single ranking. This page characterizes where each name sits — it isn't a buy or sell call on any security. It's the map we use to frame the operating conversation that follows.

    Everything on this page

    03 · SITUATION MAP Where the Money Sits: Price Against Growth for the 10 Rated Names Cut on EV / EBITDA vs the sector median (5.7x) (rows) and revenue growth vs the covered median (4%) (columns) · observations, not recommendations Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Situation boundaries are the cohort's own medians (Directional, NeuraCap view). This page characterises situations; it does not recommend buying or selling any security. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 10 Priced up and Growing Above-median multiple · above-median revenue growth 2 names ArcelorMittal S.A. (MT) · Commercial Metals Company (CMC) ArcelorMittal S.A. (MT) and Commercial Metals Company (CMC) sit above the middle of the range on price and above it on revenue growth. With two names here, read it as a position to defend rather than a pattern — mix and contract book are what a buyer would test behind it. Priced up on Steady Earnings Above-median multiple · below-median revenue growth 3 names Nucor Corporation (NUE) · Steel Dynamics, Inc. (STLD) · Cleveland-Cliffs Inc. (CLF) Nucor Corporation (NUE), Steel Dynamics, Inc. (STLD) and Cleveland-Cliffs Inc. (CLF) price above the middle with growth below it. The premium on CY2027E EBITDA sits alongside steady margin and scale rather than volume expansion, and that is the position the top of the range is holding. Growing, Priced Below Below-median multiple · above-median revenue growth 3 names Ternium S.A. (TX) · Companhia Siderúrgica Nacional (SID) · Metallus Inc. (MTUS) Ternium S.A. (TX), Companhia Siderúrgica Nacional (SID) and Metallus Inc. (MTUS) grow faster than the middle and price below it. The question for an owner in this corner is which part of the earnings stream the market is treating as temporary — spread, region or mix — because that is where the operating work goes. Below on Both Below-median multiple · below-median revenue growth 2 names POSCO Holdings Inc. (PKX) · Gerdau S.A. (GGB) POSCO Holdings Inc. (PKX) and Gerdau S.A. (GGB) sit below the middle on both measures. On two names the read is narrow, and the levers that show up in this set are cost curve position and value-added share rather than tonnes alone.

11. 11
    03 · GROWTH VS PROFITABILITY

    Clearing Both the Growth and Margin Bars Is Sitting with the Lower Price

    Maps the ten companies with both estimates on growth versus margin, showing median EV/EBITDA per quadrant.

    Cutting on both the 4% growth line and the 17% margin line, the names clearing both bars sit with the lower price — the opposite of what a simple growth-plus-margin story would predict. That's a small cohort, so we treat the pattern as directional rather than conclusive. It's still the clearest single data point in this report for why price and performance aren't tracking together here.

    Everything on this page

    03 · GROWTH VS PROFITABILITY Clearing Both the Growth and Margin Bars Is Sitting with the Lower Price Revenue growth (CY2027E, x-axis) vs EBITDA margin (CY2027E, y-axis) · 10 companies with both estimates · cuts at the covered medians (4% growth, 17% margin) · median EV/EBITDA per quadrant · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. Quadrant cuts are the covered set's medians. Quadrant medians on names with a meaningful EV/EBITDA (balanced n=2; margin-only n=3; growth-only n=3; neither n=2). Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 11 2% 4% 6% 10% 15% 20% 25% MARGIN ONLY median 7.4x BALANCED median 4.0x NEITHER median 5.6x GROWTH ONLY median 6.0x STLD NUE GGB CLF PKX SID CMC TX MT MTUS x: revenue growth (CY2027E) · y: EBITDA margin (CY2027E) HOW TO READ THIS The chart splits the 10 rated names on revenue growth and at a 17% EBITDA margin. On the two names clearing both bars — Ternium S.A. (TX) and Companhia Siderúrgica Nacional (SID) — the middle price is 4.0x, below the three names clearing growth only at 6.0x and the two clearing neither at 5.6x. Read it as an observation on a small set, not a rule: with two names in that corner, the diligence question is why the market treats those earnings as temporary. Rule of 40 (revenue growth + EBITDA margin ≥ 40%): 0 of 10 names clear it.

12. 12
    03 · THE AGENDA

    Four Operating Choices That Track Alongside Where a Steel Business Prices

    Lays out four operating questions that track alongside where a steel business prices in this set.

    We've translated the pricing pattern into four operating questions: hardening the earnings stream, moving down the cost curve, weighing build versus buy on tonnes, and timing capital against the reline calendar. These are framed as questions to resolve, not as recommendations. They're the practical version of what the valuation data on the last few pages is telling us.

    Everything on this page

    03 · THE AGENDA Four Operating Choices That Track Alongside Where a Steel Business Prices NeuraCap view · framed as the questions an owner or acquirer should resolve · observations, not recommendations Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All content on this page is Directional — NeuraCap advisory judgment grounded in the cohort data shown earlier. It does not constitute investment advice. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 12 Harden the Earnings Stream The top of the range in this set holds names with steady margin through slower growth. Contract share of the order book, coated and special bar quality mix, and a freight-defended service radius are the levers that make EBITDA per tonne less dependent on spot spread. What changes the answer: A quarter where metal spread falls and EBITDA per tonne holds. Move Down the Cost Curve Conversion cost per tonne and raw material position separate operators inside one business model. Captive or contracted scrap, pig iron, HBI and DRI supply reduce dependence on merchant prices when the cycle turns against you. What changes the answer: Conversion cost per tonne moving against the regional benchmark. Argue Build Versus Buy on Tonnes Strategic buyers in this record added capacity in regions they already served instead of building it. The same arithmetic runs in reverse for an owner: what a new line would cost, and how long permitting, power and interconnection would take, against the price of installed tonnes. What changes the answer: A permitted brownfield site with power in place coming available. Time Capital Against the Reline Calendar Reline and major maintenance timing, decarbonisation spend and available incentives sit behind how the market reads a mill's next five years. Funding capacity is what separates the moat reading from the overhang reading, and it is decided well before the spend lands. What changes the answer: A reline or conversion decision entering the next capital plan.

13. 13
    SECTION 04

    04

    Divider introducing Section 04, covering the nine recorded precedent steel transactions.

    This section turns to what buyers actually agreed to pay — nine recorded transactions running from May-2018 to October-2025. We'll walk through what these deals tell us about how the market prices consolidation, carve-outs and supply deals in steel.

    Everything on this page

    SECTION 04 04 PRECEDENT TRANSACTIONS What Buyers Agreed to Pay Across Nine Recorded Steel Transactions Nine transactions from May-2018 to Oct-2025, all recorded as announced. 04 of 06 Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 13

14. 14
    04 · DEAL CASE STUDIES

    Consolidation, Carve-Outs and Supply: What Steel Buyers Agreed to Pay

    Presents transaction case studies from the recorded precedent set, with multiples on LTM financials at announcement.

    These case studies pull from the transactions with disclosed terms, showing multiples on LTM financials at the point each deal was announced. Cleveland-Cliffs' $9.2B agreement for United States Steel is the largest disclosed value among the nine, and disclosed multiples elsewhere in the set run as high as 13.5x on a stainless long products carve-out. These deal multiples sit on a different basis than the CY2027E public multiples earlier in the report, so we're not claiming a spread between them — just showing what buyers have agreed to pay.

    Everything on this page

    04 · DEAL CASE STUDIES Consolidation, Carve-Outs and Supply: What Steel Buyers Agreed to Pay 1 of 21 transactions with disclosed terms, told as case studies · multiples on LTM financials at announcement where disclosed · the complete list is in the appendix · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 50 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 37 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. "Why the deal happened" is a NeuraCap read of the recorded evidence. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 14 Jul-2022 $2.2B ArcelorMittal acquires Companhia Siderúrgica do Pecém EV / LTM revenue n/a EV / LTM EBITDA n/a WHY THE DEAL HAPPENED A producer buying a producer in a region it already serves: the transaction suggests adding installed capacity where the alternative was the cost and timeline of a new line, including permits, power and ramp. It is one of the larger disclosed values in this record, consistent with a whole-company purchase rather than an asset package. HOW THE TARGET WAS VALUED The announced enterprise value is $2.2B, which places it in the whole-company bracket of this record rather than the small carve-out bracket. Where multiples are disclosed in the same record, steel purchases sit from 4.6x through 8.0x, with one specialty carve-out priced above that band.

15. 15
    SECTION 05

    05

    Divider introducing Section 05, on what the valuation gap means for an operator.

    With the public and deal evidence both on the table, this section turns to what the gap between the top and bottom of the range means operationally. We'll close with the questions it puts in front of an owner or acquirer over the next twelve months.

    Everything on this page

    SECTION 05 05 STRATEGIC IMPLICATIONS What the Gap Between the Two Ends Means for an Operator Operating moves that sit behind the spread between the two ends. 05 of 06 Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 15

16. 16
    05 · STRATEGIC IMPLICATIONS

    Durability of Earnings Is What the Top of This Range Sits Alongside

    Frames durability of earnings as the theme distinguishing the top of the valuation range.

    Across this set, durability of earnings is what sits alongside the top of the range rather than growth or scale on their own. That reframes the operating conversation: contract share, product mix and cost position look like the levers that hold a multiple up when the cycle turns. These are our views on the pattern in the data, not a forecast for any single name.

    Everything on this page

    05 · STRATEGIC IMPLICATIONS Durability of Earnings Is What the Top of This Range Sits Alongside NeuraCap view · the questions this data puts on the table for the next twelve months Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. This page is Directional — NeuraCap views drawn from the analysis in this report. Observations, not recommendations. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 16 FOR OWNERS Steadiness Is Showing up in the Price On CY2027E EBITDA the top of the range holds names with steady margin and modest growth. If your earnings swing with spot spread, the operating work is contract book, coated and special bar quality mix, and cost per tonne — the items that hold EBITDA per tonne when the spread moves. FOR BUYERS Growth Is Available Below the Middle Three of the 10 rated names grow faster than the middle of the set and price below it. Whether that gap reflects regional exposure, mix or a view on mid-cycle spread is the diligence question, and the answer sits in the order book and the cost curve rather than in the screen. FOR BOARDS Your Peer Set Runs the Same Routes The 10 rated names sit in the same diversified multi-route producer group, so the comparison a board faces is with operators running the same melt routes and product families. Position inside that group — cost curve, mix, contract share — is what the price differences here sit alongside.

17. 17
    SECTION 06

    06

    Divider introducing Section 06, the full comparables set, methodology and sources.

    This closing section carries the full universe behind every chart in the report — the public comparables, the complete transaction list, and how the analysis was built. It's where you go to trace any individual figure back to its source.

    Everything on this page

    SECTION 06 06 APPENDIX The Full Universe, Methodology and Sources Comparables detail behind every figure in the body, the valuation basis, and where each underlying disclosure lives. 06 of 06 Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 17

18. 18
    06 · PUBLIC COMPARABLES (1 OF 1)

    Public Comparables on EV / EBITDA (CY2027E), Grouped by Valuation Tier

    Lists all ten rated companies and the one unrated name with their EV/EBITDA CY2027E figures, grouped by valuation tier.

    This table carries all ten rated names against the 5.7x sector median, plus the one company in the set without an eligible multiple. It's the full public comparables list behind every chart earlier in the deck. Use it to check any single company's position before applying the operating questions we raised in Section 05.

    Everything on this page

    06 · PUBLIC COMPARABLES (1 OF 1) Public Comparables on EV / EBITDA (CY2027E), Grouped by Valuation Tier Teal shading marks a EV/EBITDA above the sector median (5.7x); amber marks below · 10 rated companies; 1 not rated (no eligible EV/EBITDA) · tickers link to the underlying source · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. All 10 rated rows are in this appendix and in the companion workbook, which carries the complete field set. Names without an eligible multiple are listed in the companion workbook. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 18 Company Ticker Segment EV EV/EBITDA (CY2027E) Rev growth EBITDA margin Rule of 40 PREMIUM — ≥6.7x · median 7.4x · 3 companies Steel Dynamics, Inc. STLD Diversified multi-route steel producers $37.4B 7.6x 1% 21% 21 Nucor Corporation NUE Diversified multi-route steel producers $61.7B 7.4x 2% 20% 22 Commercial Metals Company CMC Diversified multi-route steel producers $10.4B 6.7x 5% 15% 20 CORE — 4.5x–6.7x · median 5.7x · 4 companies Cleveland-Cliffs Inc. CLF Diversified multi-route steel producers $14.9B 6.5x 3% 10% 13 ArcelorMittal S.A. MT Diversified multi-route steel producers $65.2B 6.0x 5% 15% 21 Metallus Inc. MTUS Diversified multi-route steel producers $693M 5.3x 8% 9% 17 POSCO Holdings Inc. PKX Diversified multi-route steel producers $31.8B 4.7x 4% 11% 15 DISCOUNT — <4.5x · median 4.3x · 3 companies Gerdau S.A. GGB Diversified multi-route steel producers $11.5B 4.5x 3% 18% 21 Ternium S.A. TX Diversified multi-route steel producers $15.3B 4.3x 5% 19% 25 Companhia Siderúrgica Nacional SID Diversified multi-route steel producers $9.3B 3.8x 4% 27% 31

19. 19
    06 · PRECEDENT TRANSACTIONS (1 OF 2)

    All Precedent Transactions with Disclosed Terms, Newest First

    Lists precedent transactions with disclosed terms, newest first, part one of two.

    Of 58 recorded transactions, 21 carry disclosed terms, and this table lists them newest first with multiples on LTM financials at announcement. Deal values link back to the underlying filing for each transaction shown. This is the detail behind the deal patterns we highlighted in Section 04.

    Everything on this page

    06 · PRECEDENT TRANSACTIONS (1 OF 2) All Precedent Transactions with Disclosed Terms, Newest First 21 transactions with disclosed terms in this tier (58 recorded) · multiples on LTM financials at announcement where disclosed · deal values link to the underlying filing · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 50 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 37 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. 18 of 21 transactions shown; the rest are in the companion workbook. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 19 Date Transaction EV EV/LTM Rev EV/LTM EBITDA Why it matters Oct-2025 Ryerson Holding Corporation → Olympic Steel, Inc. $619M n/a n/a Ryerson Holding Corporation announced the purchase of Olympic Steel, Inc. in Oct-2025 at $619M. Both sit downstream of the mills, and the transaction suggests scale in processing and distribution, where the earnings stream is steadier than at the melt shop. Nov-2024 Intelcia Group S.A → Aços Laminados do Pará S.A. n/a n/a 8.0x Intelcia Group S.A announced the purchase of Aços Laminados do Pará S.A. in Nov-2024 at 8.0x EBITDA. That sits above the range the 10 rated names occupy today on CY2027E EBITDA, which is the kind of level long products assets attract when supply position is part of… Aug-2023 Cleveland-Cliffs Inc. → United States Steel Corporation $9.2B n/a n/a Cleveland-Cliffs Inc. announced an agreement for United States Steel Corporation in Aug-2023 at $9.2B, the largest disclosed value among the nine transactions here. Domestic consolidation of that size runs into antitrust review, which is part of why buyers in this… Jul-2022 ArcelorMittal → Companhia Siderúrgica do Pecém $2.2B n/a n/a ArcelorMittal announced the purchase of Companhia Siderúrgica do Pecém in Jul-2022 at $2.2B. Buying installed capacity in a market a producer already serves is the standard alternative to a greenfield or brownfield line, and it is the comparison strategic acquirers… Jul-2022 Health Advocate → Outokumpu Oyj (Stainless Steel Long Products Business) n/a n/a 13.5x Health Advocate announced the purchase of the Outokumpu Oyj (Stainless Steel Long Products Business) in Jul-2022 at 13.5x EBITDA. Stainless long products carry a different mix and a slower-to-replicate customer qualification profile than commodity tonnes, and the… Nov-2020 Cleveland-Cliffs Inc. → ArcelorMittal USA n/a n/a 6.6x Cleveland-Cliffs Inc. announced the purchase of ArcelorMittal USA in Nov-2020 at 6.6x EBITDA. A raw material and downstream business moving up into integrated flat-rolled melting is a supply-chain move, and it is the logic that recurs most often in this record. Mar-2020 Cleveland-Cliffs Inc. → AK Steel Holding Corporation n/a n/a 4.6x Cleveland-Cliffs Inc. announced the purchase of AK Steel Holding Corporation in Mar-2020 at 4.6x EBITDA. Announced at a weak point in the cycle, it shows how much of the negotiation in this sector turns on whose view of mid-cycle margin per tonne prevails. Aug-2018 Blue Coral Investment Holdings Pte. Ltd and Mountainpeak Investment Holdings Ltd. → Gerdau Hungria KFT Y CIA Sociedad Regular Colectiva $120M n/a n/a Blue Coral Investment Holdings Pte. Ltd and Mountainpeak Investment Holdings Ltd. announced the purchase of Gerdau Hungria KFT Y CIA Sociedad Regular Colectiva in Aug-2018 at $120M. Small cross-border carve-outs are how producers reshape a footprint without a… May-2018 Nippon Steel & Sumitomo Metal Corp → Nisshin Steel n/a n/a 5.8x Nippon Steel & Sumitomo Metal Corp announced the purchase of Nisshin Steel in May-2018 at 5.8x EBITDA. That sits close to the middle of where the 10 rated names trade today, a reminder that whole-company agreements in this sector are often struck near the screen…

20. 20
    06 · PRECEDENT TRANSACTIONS (2 OF 2)

    All Precedent Transactions with Disclosed Terms, Newest First

    Continues the list of precedent transactions with disclosed terms, newest first.

    This page completes the disclosed-terms transaction list started on the previous page. Together the two pages carry 18 of the 21 disclosed-terms transactions, with the remainder available in the companion workbook. It's the full evidence base behind the deal commentary earlier in the report.

    Everything on this page

    06 · PRECEDENT TRANSACTIONS (2 OF 2) All Precedent Transactions with Disclosed Terms, Newest First 21 transactions with disclosed terms in this tier (58 recorded) · multiples on LTM financials at announcement where disclosed · deal values link to the underlying filing · as of 2026-09-28 Source: NeuraCap analytics platform; market data and consensus estimates as of 2026-09-28; company disclosures via SEC EDGAR where linked. 50 precedent record(s) carry data-quality flags (deal value unit unresolved; divestiture roles reassigned; duplicate precedent id); figures are shown as recorded in the filing. 37 recorded transactions with neither a disclosed value nor a multiple are omitted from the precedent list and the case studies (kept in the companion workbook). Deal multiples are LTM at announcement (Definitive, from filings); they are not directly comparable to the CY2027E public basis and no spread is claimed. 18 of 21 transactions shown; the rest are in the companion workbook. Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 20 Date Transaction EV EV/LTM Rev EV/LTM EBITDA Why it matters Sep-2017 Ternium S.A. → CSA Siderúrgica do Atlântico n/a n/a 5.9x Feb-2017 Ternium S.A. → ThyssenKrupp Slab International B.V. n/a n/a 7.0x Value shown as recorded in the filing; deal value unit unresolved. Sep-2014 Nucor → Gallatin n/a n/a 6.5x Jul-2014 AK Steel → Severstal Dearborn n/a n/a 29.4x Jul-2014 Steel Dynamics → Severstal Columbus n/a n/a 6.4x May-2011 Olympic Steel → Chicago Tube & Iron n/a n/a 14.1x Dec-2007 Evraz Group S.A. → Claymont Steel Holdings, Inc. n/a n/a 9.7x Value shown as recorded in the filing; deal value unit unresolved. Aug-2007 U.S. Steel → Stelco Inc. n/a n/a 9.3x Value shown as recorded in the filing; deal value unit unresolved. Apr-2007 Ternium S.A. → Grupo IMSA S.A.B. de C.V. n/a n/a 6.9x Value shown as recorded in the filing; deal value unit unresolved.

21. 21
    06 · METHODOLOGY

    Sources, Assumptions and Data Quality

    Explains the report's sources, assumptions and data-quality treatment.

    This page sets out how the analysis was built — the valuation basis, what was excluded, and how data-quality flags were handled. We've kept the exclusion criteria visible rather than hidden, so every multiple in this deck can be checked against a plausibility test. It's the reference page for anyone who wants to see how a specific figure was derived.

    Everything on this page

    06 · METHODOLOGY Sources, Assumptions and Data Quality How this report was built, what was excluded, and where every underlying disclosure lives · as of 2026-09-28 Every figure in this report links to the record it was taken from. Where a figure has no link, the appendix names its source and the basis on which it was read. Steel Coverage | September 2026 | Confidential | Not investment advice 21 VALUATION BASIS Primary valuation basis: EV / EBITDA on CY2027E consensus (10 of 11 companies eligible; multiples failing the platform's plausibility gates are excluded, never plotted). EV / EBITDA on CY2027E is the lead convention: it is the sector-appropriate prior for Steel and it clears the coverage gate with 10 of 11 companies (91%). EV / Revenue, P / E are carried as a cross-check. The set earns: 10 of the 10 companies with a reported forward EBITDA carry a meaningful one, so the profit multiple leads and a revenue multiple would understate what the market is pricing. DATA QUALITY & EXCLUSIONS 11 records were excluded or quarantined by the platform's validation gates (unit errors, implausible ratios, ineligible securities and classification failures); the full ledger ships in the companion tables and never feeds a statistic in this report. DEFINITIVE VS DIRECTIONAL Definitive content is disclosed or consensus data passed through stated arithmetic. Directional content — tier boundaries, situation cuts, "why it matters" commentary, the agenda pages — is NeuraCap analyst judgment and is labeled as such where it appears. WHAT THIS REPORT IS NOT This report characterises a sector. It does not recommend buying, selling or holding any security, and no statement in it should be read as investment advice or a price target. WHERE THE DATA LIVES 573 registered source documents stand behind the figures in this report. Metric hyperlinks throughout the deck open the specific filing or data record; by publisher: sec.gov (572) · home.treasury.gov (1). The companion workbook beside this deck carries the complete source index.

22. 22

    Across These 10 Rated Names, the Higher Prices Sit Alongside the Steadier Earners.

    Closing slide restating that across the ten rated names, higher prices sit alongside steadier earners.

    Across these ten rated names, the higher prices sit alongside the steadier earners, not the faster growth. That's the finding we'd want a client to leave the room holding. The companion tables carry the full universe and source index for anyone who wants to trace a specific figure further.

    Everything on this page

    Across These 10 Rated Names, the Higher Prices Sit Alongside the Steadier Earners. NeuraCap AI — Steel Coverage September 2026 · Prepared by NeuraCap AI · Confidential Steel Coverage | September 2026 | Confidential | Not investment advice Sources & methodology 22

More coverage in Materials

• Bi-Weekly Update · Sep 28, 2026Steel Bi-Weekly Industry Events & M&A Update — September 14–28, 2026
• Bi-Weekly Update · Sep 28, 2026Paper and Plastic Packaging Bi-Weekly Industry Events & M&A Update — September 14–28, 2026
• Sector Report · Sep 28, 2026Paper and Plastic Packaging Sector Outlook — September 2026
• Bi-Weekly Update · Sep 28, 2026Paper and Forest Products Bi-Weekly Industry Events & M&A Update — September 14–28, 2026