Skip to main content
Homepage
Full 18-step deep-dive · ~20–30 min · a member feature.
FRESH Analysis Report
Aug 16, 2026
7 days ago · 100% complete
For AI assistants & researchers — machine-readable summary of this page

What this page is: Delvantic's full research page for ArcelorMittal S.A. (MT) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.

Our current read (analysis of 2026-08-23): Designation Low · Gem Score -51 (−100…+100 Quality+Value blend) · Quality -18 · Value -78 · Sentiment -64 (timing only, not weighted)

Page map (sections in order; each card carries a stable reference-name attribute you can cite):

  • profile-header / price-overview — company profile, live quote, market cap
  • extended-analysisthe core: three AI lens reads with findings, scores, and the analyst memo
  • future-predictions — our forward price-band predictions
  • market-narrative / ai-findings / gpt-critique — narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)
  • Members-only sections (render as login gates for anonymous readers): price-history, income-trend, key-metrics, financials (statement tables), insider-trading. The analysis above is public; the raw data tables require a free account.

More for machine readers: site briefing at /llms.txt · any ticker resolves at delvantic.com/stock/TICKER · raw inputs are public-company filings and market data (via licensed data feeds); every model, score, lens read, and prediction on this page is Delvantic's own analysis.

ArcelorMittal S.A.

MT NYSE
Basic Materials · Steel
Luxembourg, CO 1160, Luxembourg corporate.arcelormittal.com Updated Aug 15, 1:05am
Price
$73.86
Market Cap
$55.7B
Employees
125,416
Beta
1.75
Avg Volume
1,627,068
Last Dividend
$0.30
CEO
Mr. Aditya Mittal

ArcelorMittal S.A. is a prominent global leader in the steel and mining industry. This multinational corporation, headquartered in Luxembourg, primarily focuses on the production of steel for various applications. It operates across a range of industries, including automotive, construction, household appliances, and packaging. ArcelorMittal is renowned for its comprehensive production capabilities, from raw material extraction to the completion of finished steel products. The company plays a crucial role in the global supply chain and is deeply integrated into the infrastructure development of many economies. Its extensive international footprint includes facilities and operations across the Americas, Europe, Africa, and Asia, showcasing its vast reach and influence in the market. ArcelorMittal's commitment to sustainability and innovation is evident in its investment in cutting-edge technologies to enhance production efficiency and reduce environmental impact. As the largest steel producer globally, ArcelorMittal's performance often reflects broader economic trends, making it a significant player to watch in the industrial sector. The company's exploration of renewable energy solutions and green steel production underscores its pivotal role in the transition towards greener practices in heavy industry.

Runs with full report Generated: Aug 16, 2026 12:15am
Price Overview
Price at report time
$73.86
as of Aug 16, 12:23am (7d ago)
Change · Aug 16
+0.65 (+0.89%)
Day Range
$73.39 – $74.61
52-Week Range
$31.93 – $75.66
50-Day MA
$67.14
200-Day MA
$56.72
Volume
1,105,500.00
Right now · live
Log in to get the live feed
Members see the real-time price and the move since this report (over 7d).
Share Structure
Outstanding 754,041,698.00
Float 823,971,525.00
Free Float 109.3%
High free float — 109.3% of shares trade freely, ~-9.3% held by insiders/institutions
Very liquid — most shares trade freely. Low insider ownership can mean less management alignment, but makes large position sizing straightforward.
Price History (1 Year)
Last updated: Aug 16, 2026 12:27am (7d ago)
Revenue & Net Income Trend
The directional story — useful even when net income is negative.
Last updated: Aug 11, 2026 12:51pm (12d ago)
Why there are no quarterly figures for ArcelorMittal S.A.

This company does not file structured financial statements with the U.S. SEC, so quarterly figures aren't available from our filings-based data engine. Annual figures shown here come from the sources that do cover it.

Revenue
The top line — total sales before any costs or taxes are subtracted. A measure of how much business the company is doing.
Net Income
The bottom line — profit left after subtracting all expenses, interest, and taxes from revenue. Reflects accounting profitability, but includes non-cash items like depreciation, so it isn't the same as cash earned.
Operating Cash Flow
The real cash generated by the day-to-day business — selling products, paying suppliers, collecting from customers. Calculated from net income by adding back non-cash items and adjusting for timing (unpaid bills, unsold inventory). When OCF consistently lags net income, the reported profit may not be converting to real money.
Period Revenue Net Income Net Margin YoY/QoQ
Key Metrics
TD Twelve Data statement HEX SEC filing
Industry comparison last run: Aug 16, 2026 12:13am
P/E Ratio (Price per dollar of earnings)
HEX
Stock Price / EPS (Diluted)
17.97
Stock Price: $73.86
EPS (Diluted): 4.11
P/B Ratio (Price vs net asset value)
HEX
Stock Price / Book Value Per Share
1.00
Stock Price: $73.86
Total Equity: $56.54B
Shares: 766,000,000
EV/EBITDA (Total value vs operating profit)
HEX
Enterprise Value / EBITDA
9.79
Market Cap: $55.69B
Total Debt: $13.41B
Cash: $5.39B
EBITDA: $6.57B
Enterprise Value (Takeover price (cap + debt - cash))
HEX
Market Cap + Total Debt - Cash
$64.4B
Market Cap: $55.69B
Total Debt: $13.41B
Cash: $5.39B
Gross Margin (Revenue left after direct costs)
HEX
Gross Profit / Revenue
7.1%
Gross Profit: $4.38B
Revenue: $61.35B
Operating Margin (Revenue left after all operations)
HEX
Operating Income / Revenue
5.9%
Operating Income: $3.63B
Revenue: $61.35B
Net Margin (Revenue left as actual profit)
HEX
Net Income / Revenue
5.1%
Net Income: $3.15B
Revenue: $61.35B
ROE (Profit from shareholder equity)
HEX
Net Income / Total Equity
5.6%
Net Income: $3.15B
Total Equity: $56.54B
ROIC (Profit from all invested capital)
HEX
NOPAT / Invested Capital
5.1%
Operating Income: $3.63B
Tax Rate: 10.0%
Equity: $56.54B
Total Debt: $13.41B
Cash: $5.39B
Current Ratio (Can it pay short-term bills)
HEX
Current Assets / Current Liabilities
1.36
Current Assets: $30.61B
Current Liabilities: $22.52B
Debt/Equity (Leverage — debt vs equity)
HEX
Total Debt / Total Equity
0.24
Short-Term Debt: $2.74B
Long-Term Debt: $10.67B
Total Debt: $13.41B
Total Equity: $56.54B
Rev/Share (Top-line per share)
HEX
Revenue / Shares Outstanding
$80.09
Revenue: $61.35B
Shares: 766,000,000
Book Value/Share (Net assets per share)
HEX
(Total Assets - Total Liabilities) / Shares
$73.81
Total Equity: $56.54B
Shares: 766,000,000
FCF/Share (Real cash generated per share)
HEX
(Operating Cash Flow + CapEx) / Shares
$0.61
Operating CF: $4.81B
CapEx: -$4.34B
Shares: 766,000,000
CapEx is negative (outflow) — added to OCF to get FCF
Div Yield (Annual income from holding)
TD
Last Annual Dividend / Stock Price
0.4%
Last Dividend: $0.30
Stock Price: $73.86
Payout Ratio (Earnings paid out as dividends)
HEX
Dividends Paid / Net Income
17.2%
Dividends Paid: -$542.00M
Net Income: $3.15B
Industry Benchmarks
Last run: Aug 16, 2026 12:13am
Compares MT against LLM-researched typical ranges for its industry. One research call per industry, cached indefinitely — every stock in the same industry reuses the same baseline.
Income Statement (Annual)
Last updated: Aug 11, 2026 12:51pm (12d ago)
Metric 2021 2022 2023 2024 2025
Revenue $76.6B $79.8B $68.3B $62.4B $61.4B
Cost of Revenue $57.3B $67.3B $63.5B $56.7B $57.0B
Gross Profit $19.2B $12.5B $4.7B $5.8B $4.4B
Operating Expenses $2.3B $2.3B $2.4B $2.5B $748.0M
Operating Income $17.0B $10.3B $2.3B $3.3B $3.6B
Net Income $15.0B $9.3B $919.0M $1.3B $3.2B
EBITDA $19.5B $12.9B $5.0B $5.9B $6.6B
EPS $13.53 $10.21 $1.09 $1.70 $4.13
EPS (Diluted) $13.49 $10.18 $1.09 $1.69 $4.11
Balance Sheet (Annual)
Last updated: Aug 11, 2026 12:51pm (12d ago)
Metric 2021 2022 2023 2024 2025
Cash & Equivalents $4.2B $9.3B $7.7B $6.4B $5.4B
Total Current Assets $34.9B $37.1B $33.2B $29.4B $30.6B
Total Assets $90.5B $94.5B $93.9B $89.4B $97.7B
Current Liabilities $24.2B $22.4B $21.8B $21.8B $22.5B
Long-Term Debt $6.5B $9.1B $8.4B $8.8B $10.7B
Total Liabilities $39.2B $39.0B $37.8B $38.1B $41.2B
Total Equity $51.3B $55.6B $56.1B $51.3B $56.5B
Retained Earnings $36.7B $45.4B $46.3B $47.3B $49.9B
Cash Flow (Annual)
Last updated: Aug 11, 2026 12:51pm (12d ago)
Metric 2021 2022 2023 2024 2025
Operating Cash Flow $9.9B $10.2B $7.6B $4.9B $4.8B
Capital Expenditure -$3.0B -$3.5B -$4.6B -$4.4B -$4.3B
Free Cash Flow $6.9B $6.7B $3.0B $447.0M $471.0M
Acquisitions (net) -$25.0M -$939.0M -$2.5B -$184.0M $47.0M
Net Debt Issued / (Repaid)
Dividends Paid -$572.0M -$663.0M -$531.0M -$580.0M -$542.0M
Stock Buybacks
Net Change in Cash -$1.3B $5.2B -$1.9B -$815.0M -$1.5B
Growth Trends (YoY %)
Last updated: Aug 11, 2026 12:51pm (12d ago)
Metric 2022 2023 2024 2025
Revenue Growth +4.3% -14.5% -8.5% -1.7%
Gross Profit Growth -34.8% -62.2% +22.2% -24.4%
Operating Income Growth -39.5% -77.2% +41.5% +9.6%
Net Income Growth -37.8% -90.1% +45.7% +135.4%
EBITDA Growth -34.1% -61.0% +18.5% +10.6%
Dividend History (Last 20)
Last updated: Aug 11, 2026 12:51pm (12d ago)
Date Dividend Declaration Record Payment
2026-08-07 $0.15
2026-05-13 $0.15
2026-02-20 $0.15
2025-11-13 $0.28
2025-05-16 $0.28
2024-11-12 $0.25
2024-05-16 $0.25
2023-11-13 $0.22
2023-05-22 $0.22
2022-05-13 $0.38
2021-06-10 $0.30
2019-05-16 $0.20
2018-05-17 $0.10
2015-05-07 $0.51
2014-05-12 $0.51
2013-05-10 $0.51
2012-11-19 $0.48
2012-08-20 $0.48
2012-05-23 $0.48
2012-02-16 $0.48
0Company Classification 1Industry Landscape 2Company Momentum 3Forward Projection 4aDCF Valuation 4bEarnings Power Value 4cAnchored PE 4dReverse DCF 4eRevenue-Based DCF 4fAnchored P/S 4gScenario Analysis 4hDividend Discount Model 4iBook Value Analysis 4jInsider Activity 4fCash Flow Quality 4gDebt Maturity Risk 4hMacro Environment 4iSector Intelligence 4jRevenue Confidence 4kSensitivity Analysis 4lSector Demand Cycle 5AI Investigation 5bThesis Evaluation 6Valuation Synthesis
computed not applicable errored not yet run 13 computed · 6 not applicable · 1 errored · 4 not yet run
Narrative Economics
The story the market is telling about this stock — the intangible X-factor (founder mythology, cult dynamics, TAM-of-imagination) that moves price beyond what cash flows alone explain. After Shiller, Narrative Economics.
No narrative profile yet for MT — it's generated by the pipeline (market-narrative step).
AI Lens 4th lens · how AI reaches this business · 5-yr
2026-08-16
The creme is there an opportunity here? Neutral
AI is not the variable that decides ArcelorMittal — it shows up as a modest cost tailwind, a real data-centre/grid demand vector, and an under-discussed competitor for the cheap power its decarbonisation plan assumes.
Exposure of 29 with position 54 says this is an ABNB-pattern name: the scarce assets are mills, ore and permits, entrant compression scores 84, and the monetised tonne is safe. The trap is ai_margin_conversion at 41 — in a sector that lost ~9pp of gross margin in three years, AI process savings on a $57B cost base leak straight into spreads unless supply stays tight, so watch EBITDA per tonne against realised prices, not AI announcements. The overlooked observable is the industrial power contract line: if AI data-centre load pushes European electricity higher, the green-steel capex case degrades before any of it appears in an AI narrative.
54
AI Position
Mildly favorable — AI arrives as an input cost and a demand vector, not a threat
Cheap intelligence cannot make a ton of steel, so AI reaches ArcelorMittal mainly through plant-level yield and energy optimization, data-centre/grid-driven demand for structural and electrical steel, and — less comfortably — through AI power demand bidding up the electricity that decarbonised steelmaking depends on.
Exposure 29 Confidence 70 50 = neutral ⚑ fingerprint implies 70
Primary Tailwind

AI-driven electrification build-out (data centres, transformers, transmission towers, grid steel) adds a genuine incremental demand vector for structural and electrical steel at a time when the industry's baseline revenue CAGR is negative; separately, ML process control on furnaces, casters and mills lifts yield and cuts energy per tonne on a ~$57B cost base where a 1% conversion-cost saving is worth more than 2025's entire operating margin gap.

Primary Pressure

Steel is a price-taker: any productivity gain AI delivers is available to every integrated mill and Chinese exporter simultaneously, so savings get competed into price rather than banked — and AI data-centre load is a direct competitor for the cheap firm electricity that EAF and hydrogen-DRI decarbonisation economics assume.

Critical Hinge

Whether AI-era productivity shows up as sustained conversion-cost-per-tonne reduction that survives the cycle, or is handed to customers via spreads. Watch EBITDA/tonne by segment against realised steel prices, and the electricity/energy line in European operations.

Hard to Reproduce

Permitted integrated mills, captive iron ore, port and rail logistics, automotive qualification on exposed-panel and AHSS grades, and carbon allowances/CBAM positioning — none of which cheap software reproduces.

Forensic fingerprint same 11 factors for every stock · 0 unfavorable · 50 neutral · 100 favorable
Underlying Need Persistence do people still need this at all? 90
Physical steel demand is untouched by cheaper intelligence.
Cars, buildings, appliances, packaging and grids all require metal; no amount of machine intelligence substitutes a structural beam or an exposed automotive panel. The only need risk is material substitution (aluminium, composites), which is not AI-driven.
Automotive steel volumes vs aluminium share · Construction and grid steel order books · Chinese export tonnage into MT markets
relevance 70 · confidence 88
Solution Persistence will they still solve it this way? 82
Integrated and EAF steelmaking remains the way steel gets made; AI changes how it's run, not what it is.
Process route change is driven by carbon policy and energy prices, not AI. AI alters control systems, scheduling and maintenance inside the existing route rather than displacing it.
DRI/EAF project timelines and capex · Digital twin deployment across sites · Conversion cost per tonne disclosures
relevance 55 · confidence 75
Intelligence Commoditization does cheap AI power them or copy them? 62
Cheap AI powers MT's operations but is equally available to every rival mill.
Furnace optimisation, scrap-mix models and predictive maintenance are becoming vendor-supplied commodities; MT gains absolute cost benefit but little relative advantage, and in a price-taking commodity relative advantage is what pays.
Whether savings are quantified per tonne · Peer mills citing similar AI programs · Vendor vs in-house model ownership
relevance 45 · confidence 66
Responsibility Transfer are they paid to take the blame? 58
Some liability value in certified grades, but MT is paid for tonnes, not for absorbing risk.
Automotive and pressure-vessel qualification carries real accountability that a new entrant cannot assume, but this is a small part of the pricing structure and not an AI-created moat.
Automotive qualification wins · Warranty/quality claim incidence · Certified green-steel offtake terms
relevance 25 · confidence 60
Scarcity Migration do their assets get rarer or more common? 68
As analysis gets cheap, permits, ore, logistics and firm power get relatively scarcer — MT owns most of these except power.
AI commoditises the planning layer and leaves physical constraints binding, which favours asset owners; the exception is electricity, where AI demand actively competes with MT's decarbonisation path.
Industrial power contracts and pricing · Captive iron ore self-sufficiency ratio · CBAM/allowance cost per tonne
relevance 70 · confidence 70
Customer DIY Preference will customers just build it themselves? 92
No customer builds a blast furnace because software got cheap.
Backward integration into steelmaking is a capital and permitting problem, not an intelligence problem; AI does nothing to lower that barrier.
Automaker backward integration signals · Long-term offtake contract share
relevance 20 · confidence 85
AI Intermediation Position do AI agents go through them or around them? 55
Agentic sourcing could sharpen spot-market price discovery and compress distribution spreads.
Most volume moves on contracts with qualification requirements agents cannot bypass, but AI-enabled marketplaces and price transparency erode the informational rents in trading and service-centre distribution.
Spot vs contract mix by segment · Distribution/service-centre margin trend · Digital sales channel volumes
relevance 35 · confidence 58
Data Leverage does their data make AI better? 52
Decades of metallurgical and process data are useful internally but not a monetisable AI asset.
Plant sensor and heat-chemistry histories improve MT's own yield models, yet the data does not create a network effect, is not sellable, and rivals hold equivalent datasets on their own lines.
Cross-plant model transfer results · Yield and prime-rate improvement metrics · Any external data monetisation
relevance 40 · confidence 55
AI Margin Conversion do the AI savings become profit? 41
Real cost savings are likely; retaining them in a price-taking commodity is not.
With 7.1% gross margin on $61B revenue, a few points of conversion-cost efficiency is transformative in absolute terms — but industry-wide margin compression of ~9pp over three years shows how quickly such gains are passed through when supply is not constrained.
EBITDA per tonne vs realised price · SG&A as percent of revenue · Fixed cost reduction program delivery
relevance 75 · confidence 62
Revenue Unit Durability does the thing they charge for survive? 86
The monetised unit — a tonne of certified steel — is immune to intelligence deflation.
Nothing about cheaper AI reduces the tonnage required per car, building or transformer; volume and price risk are cyclical and policy-driven, not AI-driven.
Shipment volumes by segment · Electrical steel capacity ramp · Green-steel price premium realisation
relevance 65 · confidence 80
Entrant Compression how easily can newcomers copy them? 84
AI-native entrants are structurally impossible in primary steelmaking.
Entry requires multi-billion capex, permits, ore access and years of customer qualification; cheap software removes none of these, so MT's competitive set stays the existing global mills and Chinese overcapacity.
New capacity announcements globally · Chinese export volume trend · Trade measures and CBAM enforcement
relevance 60 · confidence 82

AI Lens thesis

ArcelorMittal's value proposition is almost entirely physical: extract ore, apply enormous energy, deliver certified tonnes into automotive, construction and packaging supply chains. The information-processing share of its cost base is small, so AI cannot substitute the product or disintermediate the seller — an agent can source steel but it still has to source it from someone with a mill. AI therefore enters through three narrow channels: (1) operating cost — predictive maintenance, furnace and blast optimisation, scrap-mix and yield models against a cost base of roughly $57B, where even modest gains dwarf net income at 2025's 7.1% gross margin; (2) demand — AI infrastructure needs structural steel, rebar, grid steel and electrical steel, a real offset to a -2.5% industry revenue CAGR; (3) input cost — the same AI build-out competes for the low-carbon electricity that green-steel capex assumes, potentially raising the hurdle on DRI/EAF projects. The commodity structure means channel (1) leaks to customers; channels (2) and (3) partially net out. Net: mildly positive, low magnitude, and swamped by cycle and Chinese export volumes.

Thesis breaker If EBITDA/tonne in Europe and NAFTA improves for consecutive years while realised prices are flat or falling, AI-era operating leverage is being retained and the read is too conservative. Conversely, a green-steel project deferred explicitly on power price or availability would confirm the electricity-competition downside.
What the market may be underestimating

Upside Electrical steel and grid-infrastructure steel — transformers, motors, transmission — is a structurally tighter, higher-spec niche than commodity flats, and AI/electrification demand raises the relative value of MT's capacity and qualification there rather than its tonnage generally.

Downside Decarbonisation capex was underwritten on assumptions of cheap, abundant renewable electricity and hydrogen; AI data-centre load competing for the same power raises the marginal cost of the very transition MT is spending billions on, and can strand or delay projects without ever appearing in an 'AI risk' disclosure.

Outcome range spread 28

39Bear case
54Central case
67Bull case
Three headline numbers, deliberately never blended: Position (which way), Exposure (how much it matters at all), Confidence (how sure). The fingerprint asks every stock the same 11 questions so companies a sector label would lump together get told apart. Not an input to GEM/Coal or the Q/V/S lenses.
Growth Outlook
Analyzed 2026-08-17 16:20

The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.

Holding Earnings are inflecting hard off a cyclical trough (+135% YoY) while the top line still shrinks slightly — a profit recovery, not a growth story, inside a steel category that is only steady. conf 6/10
Cyclical Category growing · Steel category median recent growth is +3.3% (against a -2.5% long-term CAGR — i.e., a cyclical bounce off a declining base). MT's revenue is -1.7% YoY, so it is lagging the category's recent turn on revenue while outpacing it on earnings recovery. That mix says regional exposure (Europe-heavy, import-pressured) rather than product-level customer defection.
Next 2 quarters
Growing
Favorable earnings comparables, the Calvert/India volume ramp landing in the print, and tightening EU import quotas should keep the earnings recovery visible for two more quarters even with flat-to-soft revenue.
≈ inline with expectations
Year 1
Growing
Full-year earnings should rise on trough comparables, self-help cost programs and new-capacity volumes, but revenue growth stays low-single-digit at best given Chinese-export price pressure and soft European construction.
≈ inline with expectations
Years 2–3
Holding
Structurally, earnings power settles at a mid-cycle level: new Indian and US capacity adds volume, but permanently lower category margins (-9pp industry-wide), decarbonization capex, and the Chinese supply overhang cap the through-cycle spread. Revenue growth is unlikely to break above GDP-ish.
↓ below expectations
The creme: each rung's call measured against what's already printed (vs analyst estimates · vs guidance / FY consensus · vs price-implied growth) — expectations in print are already in the price, so only the variant margin can pay. Hover a rung's chip for the margin read.
Growth drivers
62 Earnings inflection off trough — Recent earnings YoY +135% with 8.5x multi-year earnings CAGR base effect: spreads, cost-out and mix are recovering from a genuine profit trough even as tonnage/pricing stay soft. Operating leverage in steel is violent in both directions, and the direction of change is currently up.
45 Structural capex pipeline adds volume, not just price — Company-specific growth units — Calvert/US EAF ramp, AM/NS India expansion toward higher nameplate, downstream/renewables plate — are self-funded volume additions that grow revenue independent of steel price. These are contracted assets already under construction, so the volume shows up regardless of cycle timing.
41 European trade protection tightening — CBAM phase-in and tighter EU safeguard quotas mechanically restrict the import share that has been suppressing European realized prices — MT's most margin-levered region. This is a policy-driven floor under pricing rather than a demand call.
27 Per-share compounding via buyback — Persistent share-count reduction converts flat absolute EBITDA into rising per-share earnings; four consecutive EPS beats (+1%, +24%, +6%, and a large loss-narrowing) are partly this mechanism plus cost discipline.
Growth risks
69 Chinese export overhang caps pricing — Record-scale Chinese steel exports remain the single binding constraint on global spreads. No MT self-help fixes it; it caps how far the earnings recovery can run and is why revenue CAGR is -5.2% despite volume investment.
55 Industry-wide margin compression — Category gross/operating/net margins down ~9.5/9.2/7.1pp over three years and industry earnings CAGR -30.4%. MT's recovery is off a low base into a structurally lower-margin regime, not back to prior peak economics.
48 Top line still contracting — Recent revenue YoY -1.7% versus category median recent growth +3.3% — the company is not participating in the modest category upturn on the revenue line, which is the clearest signal that volume/mix in Europe remains impaired.
39 Macro/demand headwinds on end markets — 10y at 4.63 with a flat-ish curve pressures construction starts and auto affordability — MT's two largest demand pools. A European industrial slowdown would strand the fixed-cost base quickly.
29 Green-steel capex without paid premium — Decarbonization spend is committed while customer willingness to pay a green premium is unproven; that combination compresses returns on the very capex meant to drive growth.
Steel remains policy- and geography-determined, not technology-disrupted. Two forces dominate the next 24 months: Chinese overcapacity exporting deflation into every open market, and Western trade/carbon walls (EU CBAM, US Section 232-style tariffs) rebuilding regional price islands. MT is unusually well placed for that bifurcation — it owns capacity inside both protected blocs plus low-cost Indian growth — but the same walls raise its own decarbonization capex bill. Energy-transition demand (grid steel, wind plate, EV silicon steel) is real but too small yet to offset flat construction. Higher-for-longer rates keep the construction demand pool subdued. Net: a business whose earnings power is recovering toward a mid-cycle level that is structurally below the 2021-22 peak.
Growth position composite -20
ShrinkingStallingHoldingGrowingAccelerating
70Next 2 quarters · Growing
70Year 1 · Growing
50Years 2–3 · Holding
-20Composite (−100…+100)
A research prediction, not advice. Forward-graded: each rung is scored against the prints that follow it. Not an input to the GEM designation — track record first.
Claude Reading
Independent analyst synthesis · claude-opus-4-7 · generated 2026-08-16 00:25:37
Verdict Fairly valued around $74 with asymmetric upside — synthesis's $36 fair value is a DCF artifact of trough FCF and elevated transition capex; real range is $60–$90 depending on China supply discipline.

Looking at the raw numbers first: revenue has bled from $79.8B (2022) to $61.4B (2025), a 23% peak-to-trough decline, while net income collapsed from $14.96B (2021) to $3.15B (2025) — roughly a 79% haircut. Gross margins compressed from 25.1% (2021) to 7.1% (2025), and operating margins from 22.2% to 5.9%. Yet the 2025 result is a *recovery* from 2024's $1.34B in NI, so we're seeing an inflection off the trough, not a fresh collapse. FCF at $471M against $4.34B capex means capital return capacity is minimal — dividend is 0.4% yield, payout only 17%, so no imminent cut, but no juice either. Balance sheet is fine: $13.4B debt vs $56.5B equity, D/E of 0.24, current ratio 1.36. The "high debt/interest coverage" flag from the synthesis looks overstated to me — MT has been an aggressive deleverager for a decade.

Now the models. The synthesis pins fair value at $36–$41 and calls the stock 51% overvalued. I think this is wrong in *direction* and probably wrong in *magnitude*. Here's why: a $36 fair value implies steady-state earnings of roughly $2B on 18x, or $3B on 12x. MT already earned $3.15B in 2025 on what everyone acknowledges are trough conditions. Mid-cycle earnings for MT historically run $5–7B; even conservatively normalize to $4.5B and put a cyclical 10x on it and you get $45B market cap — below current $55.7B, but not by 50%. Put 12x on $5B and you're at $60B, right around spot. The DCF-driven $36 anchor is being computed off suppressed FCF ($471M) that is depressed by elevated decarbonization capex ($4.34B, well above maintenance) — that's a modeling artifact, not intrinsic value destruction. The P/B of 1.0x on $56.5B of tangible-heavy equity is the real anchor here, and it says the market is *not* paying a narrative premium; it's paying book.

The contrarian case against my read: Chinese overcapacity is genuinely structural (thesis eval top bear at 75 weight is right to flag it), and MT's ROIC of 5.1% and ROE of 5.6% are terrible even for a cyclical mid-point — if this *is* mid-cycle rather than trough, then the stock deserves to trade at book and the $36 DCF might actually be generous. The 85% earnings CAGR is a base-effect illusion off 2023's $919M nadir. Revenue is still declining YoY (-1.7%), which is inconsistent with a "trough behind us" story. And the "green steel" narrative genuinely is aspirational — European carbon border adjustment (CBAM) enforcement has been repeatedly weakened. If steel prices roll over in 2026 on China property drag continuing, earnings could easily halve again and the stock retests $50s.

Where I land: the synthesis and market forces engine are too bearish because they're extrapolating trough FCF and treating decarbonization capex as permanent cash destruction rather than a discretionary investment cycle. The narrative layer's assessment that "$74 embeds cyclical peak pricing" is inverted — 5.1% net margins and $3.15B NI are not peak-cycle numbers, they're low-mid cycle. The stock at 1.0x book, 0.9x sales, and 9.8x EV/EBITDA is priced for a modestly recovering cyclical, not a euphoric one. I dissent from the "overvalued to $36" verdict. My read is fairly valued to modestly undervalued in the $70–$85 range, with genuine downside to $50–55 if China dumps steel again and genuine upside to $95–100 if 2026 brings even a mild cyclical recovery + CBAM enforcement. This is a coin flip weighted slightly toward the upside because you're paying book value for the world's largest ex-China steelmaker with a deleveraged balance sheet. Not table-pounding, but the models here are anchoring on the wrong number.

GPT Reading
Independent reading · gpt-5.4 · generated 2026-08-16 00:25:51
Verdict Overvalued at $73.86 — fair value looks closer to $45-$55 unless free cash flow recovers above $2.5B and operating margins prove they can stay above 7%.

What jumps out first is that the current valuation is not obviously “deep value” on the actual earnings and cash flow on offer today. At $55.7B of market cap and about $63.7B of enterprise value after netting $5.4B of cash against $13.4B of debt, ArcelorMittal is trading at roughly 18x 2025 earnings, about 9.8x EV/EBITDA, 1.0x book, and nearly 0.9x sales. Those multiples might be fine for a decent industrial with durable mid-teens returns; they are not cheap for a steelmaker generating just 5.6% ROE and 5.1% ROIC in the latest year. The income statement tells a very cyclical, low-quality story: revenue fell from $79.8B in 2022 to $61.4B in 2025, while operating income dropped from $10.3B to $3.6B. Yes, net income recovered from $919M in 2023 to $3.15B in 2025, but that recovery still leaves earnings only about one-fifth of the 2021 peak and one-third of the 2022 level. Paying book value for a commodity producer only makes sense if that book can earn materially above cost of capital; the latest returns do not show that.

The bigger problem is cash generation. 2025 operating cash flow was $4.81B, but capex consumed $4.34B, leaving just $471M of free cash flow. On the current market cap, that is an FCF yield under 1%. For a cyclical steel business, that is a severe mismatch. Bulls will argue capex is temporarily elevated and should support higher future profitability, but until that is visible, the stock is being priced on hoped-for normalization rather than delivered owner earnings. Even giving credit for a solid balance sheet relative to the industry—net debt is only about $8.0B and debt/equity is 0.24—the equity still looks expensive because the balance sheet strength is already capitalized. This is not a distressed steel company mispriced at half of liquidation value; it is a financially sound steel company being valued as though mid-cycle earnings power is comfortably above what 2025 demonstrates.

The raw operating progression also undermines the idea that the market is simply looking through a trough. Gross margin was 25.1% in 2021, 15.7% in 2022, then just 6.9% in 2023, 9.3% in 2024, and 7.1% in 2025. That is not a clean rebound path; it is a business still struggling to convert revenue into profit in a harsher cost and pricing environment. Operating margin improved from 3.4% in 2024 to 5.9% in 2025, but gross profit actually fell from $5.79B to $4.38B despite only a modest revenue decline, implying the net income improvement likely benefited from below-the-line factors or cost actions that may not be durable. I do not buy the simplistic “P/E is high because earnings are depressed” excuse unless I can also see compelling evidence that normalized earnings are much higher. The last three years instead suggest normalized earnings may be far closer to $2B-$4B than to the $9B-$15B earned in the post-pandemic windfall.

The best argument against my skepticism is straightforward: this company has survived the downshift without balance-sheet stress, and if steel spreads improve even modestly, earnings torque is large. At 1.0x book and only 0.92x sales, the shares are not optically frothy. Net debt of about $8B is manageable versus $56.5B of equity, and 2025 net income more than doubled year over year from $1.34B to $3.15B despite lower revenue, showing management can protect the bottom line. If 2025 capex is unusually heavy, then free cash flow could snap back sharply; a move from $0.5B to even $3B-$4B of FCF would transform the valuation optics. A smart bull would also note that 2021-2022 proved this asset base can throw off extraordinary cash in tighter steel markets, so the stock has embedded optionality to any supply discipline, China stimulus, trade protection, or industrial demand surge. I weigh that less heavily because commodity optionality is not the same as franchise value, and the market is already paying a full price for that option.

What would change my mind is evidence that 2025 is indeed a cash-flow trough rather than the new mid-cycle. Concretely, I would want to see operating margin hold above 7% and free cash flow move sustainably above $2.5B without levering up the balance sheet; that would put the stock on a mid-single-digit FCF yield and support a valuation closer to today’s price. I would also revisit the bear case if ROIC climbs back toward 8%-10% while capex moderates from $4.34B, because then book value would deserve a premium rather than merely par. Until those numbers show up, I think the stock is priced as if the recovery is already established when the reported data still looks like a low-return cyclical struggling to earn its keep.

Grok Reading
Independent reading · grok-4.5 · generated 2026-08-16 00:26:42
Verdict Overvalued at $73.86; mid-cycle fair value nearer $45–55 on ~$4B NI and sub-par ROE at 1× book

ArcelorMittal’s financials describe a classic late-cycle industrial that has already surrendered most of its peak earnings power and is now being asked to fund a multi-year capital reset on thin free cash flow. Revenue has slid from $79.84B in 2022 to $61.35B in 2025, a cumulative contraction of roughly 23 percent, while net income—though sharply higher year-over-year at $3.15B versus $1.34B in 2024 and $919M in 2023—remains barely one-fifth of the $14.96B earned in 2021. Gross margin has compressed from the mid-20s to 7.1 percent and operating margin sits at 5.9 percent; the resulting ROE of 5.6 percent and ROIC of 5.1 percent are below any reasonable cost of capital for a global steelmaker. Operating cash flow of $4.81B looks adequate until capex of $4.34B leaves free cash flow of only $471M, a sub-1 percent FCF yield on the $55.7B equity value. At 18× trailing earnings, 1.0× book and 9.8× EV/EBITDA, the market is capitalizing a still-depressed but recovering profit base as if mid-cycle returns and cash conversion are already secured. That is the core mispricing: investors are paying peak-cycle multiples for trough-to-mid earnings while the balance-sheet capacity is being absorbed by heavy reinvestment rather than returned to shareholders.

The prior models correctly flag the narrative premium—roughly a doubling of a mid-cycle DCF near $36–41—but they overstate balance-sheet fragility. Net debt is only about $8B against $56.5B of equity (debt-to-equity 0.24), so the “high debt risk” label is not supported by the numbers; liquidity and leverage are fine. The real vulnerability is cash conversion and return on capital, not solvency. Earnings CAGR of 85 percent and the 135 percent jump in recent NI are optically impressive yet rest on an extremely low base and have not translated into free cash; revenue CAGR of –5.2 percent and FCF CAGR of –60.6 percent are the more durable signals. Chinese overcapacity remains structural, European energy costs have permanently lifted the cost curve, and the green-steel capex program is diluting near-term returns without yet producing priced-in pricing power. On normalized mid-cycle net income of roughly $3.5–4.5B, a 10–12× multiple and a 0.7–0.85× book multiple both point to equity value in the mid-$40s to low-$50s, well below the $73.86 print.

The strongest opposing case is straightforward and must be weighed. Steel equities are operating-leverage machines: a $50–100 per tonne lift in realized prices can double earnings inside a year, and MT already demonstrated that path from 2023 to 2025. Trading at exactly book value with a fortress balance sheet historically marks the zone where patient capital accumulates, not where it exits; true cycle bottoms have printed closer to 0.5–0.6× tangible book. If China stimulus tightens scrap and iron-ore balances, or if carbon-border adjustments actually ration Chinese imports into Europe, MT’s scale and partially decarbonized asset base could sustain mid-cycle ROE closer to 10–12 percent rather than the current 5–6 percent. In that scenario the 18× multiple on recovering earnings becomes mid-cycle 8–10× forward, and $74 is no longer expensive. The bull also notes that P/S of 0.92× embeds little revenue growth optimism, so the multiple risk is concentrated in margins and FCF, not top-line disappointment.

That upside path is real but is already the embedded bet, not a free option. I therefore stay on the side of overvaluation: the stock discounts a durable margin recovery and FCF inflection that the last three years of data have not yet delivered. What would flip the view is sustained free cash flow above $2.5–3B (implying capex discipline or volume recovery), gross margin holding above 11–12 percent for four consecutive quarters, and clear evidence that Chinese net exports are structurally declining rather than merely pausing. Absent those prints, $73.86 remains a rich price for a 5 percent ROIC steel business.

Big-3 Panel — where each AI stands
Each AI above independently stated a direction (undervalued, fairly valued, or overvalued) and how strongly it believes it (conviction, 0–5). We combine those into a Bull-Bear Index on a 0–10 scale: 5 is neutral, 10 is maximum bullish (undervalued at full conviction), 0 is maximum bearish. We compute the score ourselves with the same arithmetic for every seat — the models never grade their own bullishness — so the three are directly comparable. Δ shows how far each seat sits from the panel average of 2.3; a large Δ marks the dissenting voice, usually the one worth reading.
Claude claude-opus-4-7 5.0
fairly valued · conviction 3/5 · Δ +2.7 vs panel · self: 6.0
GPT gpt-5.4 1.0
overvalued · conviction 4/5 · Δ -1.3 vs panel · self: 3.0
Grok grok-4.5 1.0
overvalued · conviction 4/5 · Δ -1.3 vs panel · self: 3.0
Advanced Analysis Forensic deep-dive · separate lenses
Separate reads — Company Quality (is it a great business?), Valuation (is it mispriced?), and General Sentiment (how macro + narrative are pushing it), plus AI Impact (how the AI wave reshapes it), kept deliberately apart · 2026-08-16 00:35:28
Delvantic - Cairn AI
Pass on price — revisit low $40s 8/10
A mid-quality cyclical trading at roughly 2x a defensible fair value while momentum quietly rolls over — this is a pass on price, not a buy.
The cruxWhether current $73.86 is capitalizing a durable green-steel/AI-infrastructure demand step-up or a fading late-cycle bid on collapsing FCF — and the cash flow math (471M FCF vs 3.15B NI) says it's the latter.
Forensic checks Derived mechanically from MT's filed financials — not from the AI lenses
Earnings QualityHigh Earnings Quality
The four lensesswitch a tab for its full read — score + evidence
Company Quality
-18
Mixed
edge √Σ 89 · risk √Σ 107 · conf 6/10

ArcelorMittal is a classic mature cyclical caught on the wrong side of the steel cycle. Revenue has fallen every year since the 2022 peak ($79.8B to $61.4B), gross margin has compressed from 25.1% in 2021 to 7.1% in 2025, and net income has cratered from $14.96B to $3.15B. Operating margin at 5.9% is a shadow of the 22.2% posted in 2021. This is the shape of a commodity trough, not a broken business, but there is nothing in the trajectory that suggests durable pricing power or a widening moat. Earnings integrity is a genuine strength: accruals -1.6% of assets, OCF/NI of 3.05x, Beneish M at -2.41 all indicate the reported profits are real and, if anything, conservatively stated. Share count has been steadily reduced from 1.11B to 766M (-31% over five years), which is real per-share value creation and disciplined capital return. Altman Z at 2.38 sits in the grey zone - not distress, but not fortress either, consistent with a capital-intensive cyclical. The soft spot is cash conversion in the down-cycle. FCF has collapsed from $6.90B in 2021 to $471M in 2025 despite $3.15B of net income, meaning current FCF/NI is only ~15% - the opposite of the OCF/NI signal and a sign that working capital and capex are eating the cycle's meager earnings. For a heavy-industrial with grey-zone solvency, that thin FCF cushion in a trough year is the main quality concern.

Strengths 2
m70
Aggressive buyback discipline
Diluted shares reduced from 1.11B (2021) to 766M (2025), a 31% reduction, materially protecting per-share value through the downcycle.
m55
Clean earnings quality
Accruals -1.6% of assets, OCF/NI 3.05x, Beneish M -2.41 - no signs of earnings management; reported numbers appear conservative.
Concerns 4
m65
Severe margin compression
Gross margin collapsed from 25.1% (2021) to 7.1% (2025); operating margin from 22.2% to 5.9%. Consistent with commodity cycle but shows no pricing durability.
m60
FCF has thinned dramatically
FCF fell from $6.90B (2021) to $471M (2025) - only ~15% of $3.15B net income - suggesting working capital or capex is absorbing most of the cash despite headline profits.
m40
Altman Z in grey zone
Z-score 2.38 places solvency in the grey band - not distressed but not fortress; matters more given capital intensity and thin trough FCF.
m45
Structurally cyclical, no visible moat
Revenue down four consecutive years ($79.8B to $61.4B) and margin swings of ~1800bps show the business is a price-taker in a global commodity.
This is a well-managed cyclical, not a great business. Management is doing the right things - honest books, aggressive share retirement, no leverage panic - but the underlying economics are commodity steel, and the 2021 numbers were a cycle peak that will not recur soon. The FCF collapse from $6.9B to $471M in four years while net income stayed positive is the number that keeps me from grading higher; it says the current earnings are barely converting to distributable cash. Solid, not strong.
Verify before trusting this (5)
  • Working capital movements and capex intensity in 2024-2025 explaining the NI-to-FCF gap
  • Debt maturity schedule and net leverage vs EBITDA at trough
  • Whether buybacks were funded from cash or incremental debt
  • Segment/geographic exposure and any evidence of pricing power in specialty products
  • Pension and environmental liabilities not visible in headline balance sheet
Valuation / Mispricing
-78
Rich
edge √Σ 20 · risk √Σ 126 · conf 7/10
price $73.86 vs deserved ~$36-41, roughly 45-50% overvalued on cycle-adjusted math. attractive below $40.00

The composite fair value sits at $40.68 and the signal-adjusted FV at $35.96, implying about -51% downside from the $73.86 print. The DCF ($13.02) and EPV floor ($25.56) both scream that the through-cycle earnings power of a commodity steelmaker does not support today's price; only the anchored-PE method ($111.11) points the other way, and that is exactly the method most easily fooled by peak-cycle EPS being extrapolated - it should be discounted heavily here. Company-quality is Mixed (score -18) with FCF having collapsed from $6.9B to $471M in four years while net income stayed positive, which is the classic tell of a cyclical top being capitalized as if it were normal earnings. A reasonable deserved value blending EPV floor with a modest cyclical premium lands in the $30-45 zone. Against $73.86, that is a rich-to-overvalued setup: buyers here are paying a ~2x premium to EPV and effectively underwriting that current spread economics persist. The bull case (green steel, structural demand, consolidation) is real but aspirational and largely priced in; the bear case (commodity mean reversion, peak-EPS optics) has the math on its side. Honest read: not a fat short, but not a value setup either - this is late-cycle pricing on a mid-quality cyclical.

Cheap signals 1
m20
Honest accounting and buybacks
High earnings-quality signal and aggressive share retirement modestly raise deserved value versus a typical commodity peer, but not enough to close a 50% gap.
Rich / priced-in 5
m72
Price ~2x composite FV
$73.86 vs composite $40.68 and signal-adjusted $35.96 - implied downside of -51% on the blended math.
m68
EPV floor at $25.56
Even the earnings-power floor sits 65% below the market price, meaning today's quote requires sustained above-normal earnings, not just steady-state.
m55
FCF collapse ignored
FCF fell from $6.9B (2021) to $471M in four years while price stays elevated - market is capitalizing net income that cash flow no longer supports.
m45
Late-cycle anchored PE is the outlier
Anchored-PE FV of $111 relies on recent peak EPS; for a commodity cyclical this method systematically overstates deserved value at the top.
m30
DCF at $13 flags method risk both ways
A DCF 5x below price suggests the model is punishing on terminal assumptions; even taking it as a soft floor rather than gospel, the EPV corroborates the rich read.
I can't call this cheap at $73.86 when the EPV floor is $25 and cash flow has evaporated. This is a mid-quality cyclical being priced like a structural winner on the green-steel narrative. I want it in the low $40s or below before valuation gets interesting; today it's a pass on price, not a short with conviction but certainly not a buy.
Verify before trusting this (5)
  • Normalized mid-cycle EBITDA per ton assumption vs latest guidance
  • Sustainable capex for decarbonization and its drag on FCF
  • Working capital swings distorting recent FCF prints
  • Segment mix - how much of EBITDA is from higher-quality NAFTA/Europe flat vs commodity long products
  • Any one-off impairments or tax items inflating recent EPS
General Sentiment
-64
Headwind
tail √Σ 35 · head √Σ 110 · conf 6/10

The macro tape is mildly risk-on with VIX at 14 and the S&P near highs, which normally amplifies a beta-1.75 name to the upside. But the pressure landing on MT specifically is not clean: the active narrative is cyclical-late-stage with fragile durability and low cult coefficient, meaning there is no true-believer base to defend the story on any wobble. Steel is exactly the archetype that gets sold first when the growth story cracks, and rates at 4.63% with a stretched market PE add a slow-drip headwind to capital-intensive commodity names. Momentum is already rolling over (-5.2% CAGR, cash generation weakening), which tells you the tape is already marking this down even while the broader index holds up. Analyst tone and news flow are quiet - no fresh upgrades, no green-steel catalyst, no infrastructure headline to reignite the bull story. In a fragile narrative regime, silence is a headwind because the story needs feeding to hold a premium multiple. Net: the risk-on backdrop is a modest tailwind, but the stock-specific forces (fading narrative, weakening momentum, late-cycle archetype, high beta into any macro wobble) outweigh it.

Tailwinds 1
m35
Risk-on tape lifts high-beta names
VIX 14 and index near highs is a supportive backdrop, and beta 1.75 means MT should get amplified upside if the tape holds. But it is a general lift, not a MT-specific bid.
Headwinds 5
m60
Fragile late-cycle narrative with no cult support
Cyclical-late-stage archetype with fragile durability and low cult coefficient means the story cannot absorb bad prints. Steel narratives crack fast when growth doubts appear, and there is no loyal holder base to defend the multiple.
m55
Momentum already rolling over
Negative CAGR and weakening cash generation signal the tape is quietly de-rating this name even in a benign macro. Trend is the tell that sentiment is leaking regardless of index levels.
m40
Rates and stretched market PE press capital-intensive cyclicals
10y at 4.63% and market PE 26 is a background headwind for a high-capex, low-ROIC commodity producer. Not decisive on its own but adds to the drag.
m30
Silent news flow on a story that needs feeding
No green-steel, infrastructure, or M&A catalysts in the tape. A fragile narrative decays without fresh proof points, and analyst tone is not stepping up to defend it.
m55
High beta cuts both ways into any macro wobble
Beta 1.75 means any risk-off flinch (a hot CPI, a curve shift, a China demand scare) hits MT roughly twice as hard as the index. With the regime only 10 days old and medium confidence, that asymmetry is a real overhang.
Net pressure leans headwind. The risk-on tape is a genuine but modest lift, and a beta-1.75 name should be enjoying it - but it is not, because the stock-specific narrative is late-cycle, fragile, and has no cult to defend it, and momentum is already fading. That divergence (benign macro, weak tape on this name) is the signature of sentiment quietly leaking out of the story. Not a strong headwind - no active narrative collapse, no downgrade cascade - but a persistent, ordinary crosswind pushing down more than the market lifts up.
Verify before trusting this (5)
  • Any China stimulus or infrastructure headline that could reignite the steel demand narrative
  • Sell-side target revisions or a downgrade cluster that would confirm fading tone
  • Steel spot prices and spreads - the real-time tell for whether late-cycle pricing power is holding
  • A shift in the risk regime (VIX break above 18 or 10y above 4.8%) that would punish high-beta cyclicals disproportionately
  • Green-steel policy or customer commitment news that could refresh narrative durability
The market-wide tape + this name's exposure to it (beta / sector / narrative durability). Context on the non-fundamental pressure — not a call on the business or the price. processId: detail-general-sentiment
AI Impact
+40
Mildly favorable — AI arrives as an input cost and a demand vector, not a threat
opp √Σ 96 · thr √Σ 0 · conf 7/10

ArcelorMittal's value proposition is almost entirely physical: extract ore, apply enormous energy, deliver certified tonnes into automotive, construction and packaging supply chains. The information-processing share of its cost base is small, so AI cannot substitute the product or disintermediate the seller — an agent can source steel but it still has to source it from someone with a mill. AI therefore enters through three narrow channels: (1) operating cost — predictive maintenance, furnace and blast optimisation, scrap-mix and yield models against a cost base of roughly $57B, where even modest gains dwarf net income at 2025's 7.1% gross margin; (2) demand — AI infrastructure needs structural steel, rebar, grid steel and electrical steel, a real offset to a -2.5% industry revenue CAGR; (3) input cost — the same AI build-out competes for the low-carbon electricity that green-steel capex assumes, potentially raising the hurdle on DRI/EAF projects. The commodity structure means channel (1) leaks to customers; channels (2) and (3) partially net out. Net: mildly positive, low magnitude, and swamped by cycle and Chinese export volumes.

AI opportunities 7
m56
Underlying Need Persistence
Physical steel demand is untouched by cheaper intelligence.
m35
Solution Persistence
Integrated and EAF steelmaking remains the way steel gets made; AI changes how it's run, not what it is.
m11
Intelligence Commoditization
Cheap AI powers MT's operations but is equally available to every rival mill.
m25
Scarcity Migration
As analysis gets cheap, permits, ore, logistics and firm power get relatively scarcer — MT owns most of these except power.
m17
Customer DIY Preference
No customer builds a blast furnace because software got cheap.
m47
Revenue Unit Durability
The monetised unit — a tonne of certified steel — is immune to intelligence deflation.
m41
Entrant Compression
AI-native entrants are structurally impossible in primary steelmaking.
AI threats 0

None surfaced.

AI is not the variable that decides ArcelorMittal — it shows up as a modest cost tailwind, a real data-centre/grid demand vector, and an under-discussed competitor for the cheap power its decarbonisation plan assumes. Exposure of 29 with position 54 says this is an ABNB-pattern name: the scarce assets are mills, ore and permits, entrant compression scores 84, and the monetised tonne is safe. The trap is ai_margin_conversion at 41 — in a sector that lost ~9pp of gross margin in three years, AI process savings on a $57B cost base leak straight into spreads unless supply stays tight, so watch EBITDA per tonne against realised prices, not AI announcements. The overlooked observable is the industrial power contract line: if AI data-centre load pushes European electricity higher, the green-steel capex case degrades before any of it appears in an AI narrative.
Verify before trusting this (8)
  • EBITDA per tonne vs realised price
  • SG&A as percent of revenue
  • Fixed cost reduction program delivery
  • Automotive steel volumes vs aluminium share
  • Construction and grid steel order books
  • Chinese export tonnage into MT markets
  • Industrial power contracts and pricing
  • Captive iron ore self-sufficiency ratio
The structural effect of the AI wave on this specific business over the next ~5 years — demand, cost leverage, moat, barriers to entry, position in the AI stack. The reality beneath the AI story, not the story's market pressure (General Sentiment owns that) — and not a call on the business today or the price.
Growth Outlook
-20
Holding
edge √Σ 91 · risk √Σ 112 · conf 6/10

Steel remains policy- and geography-determined, not technology-disrupted. Two forces dominate the next 24 months: Chinese overcapacity exporting deflation into every open market, and Western trade/carbon walls (EU CBAM, US Section 232-style tariffs) rebuilding regional price islands. MT is unusually well placed for that bifurcation — it owns capacity inside both protected blocs plus low-cost Indian growth — but the same walls raise its own decarbonization capex bill. Energy-transition demand (grid steel, wind plate, EV silicon steel) is real but too small yet to offset flat construction. Higher-for-longer rates keep the construction demand pool subdued. Net: a business whose earnings power is recovering toward a mid-cycle level that is structurally below the 2021-22 peak.

Growth drivers 4
m62
Earnings inflection off trough
Recent earnings YoY +135% with 8.5x multi-year earnings CAGR base effect: spreads, cost-out and mix are recovering from a genuine profit trough even as tonnage/pricing stay soft. Operating leverage in steel is violent in both directions, and the direction of change is currently up.
m45
Structural capex pipeline adds volume, not just price
Company-specific growth units — Calvert/US EAF ramp, AM/NS India expansion toward higher nameplate, downstream/renewables plate — are self-funded volume additions that grow revenue independent of steel price. These are contracted assets already under construction, so the volume shows up regardless of cycle timing.
m41
European trade protection tightening
CBAM phase-in and tighter EU safeguard quotas mechanically restrict the import share that has been suppressing European realized prices — MT's most margin-levered region. This is a policy-driven floor under pricing rather than a demand call.
m27
Per-share compounding via buyback
Persistent share-count reduction converts flat absolute EBITDA into rising per-share earnings; four consecutive EPS beats (+1%, +24%, +6%, and a large loss-narrowing) are partly this mechanism plus cost discipline.
Growth risks 5
m69
Chinese export overhang caps pricing
Record-scale Chinese steel exports remain the single binding constraint on global spreads. No MT self-help fixes it; it caps how far the earnings recovery can run and is why revenue CAGR is -5.2% despite volume investment.
m55
Industry-wide margin compression
Category gross/operating/net margins down ~9.5/9.2/7.1pp over three years and industry earnings CAGR -30.4%. MT's recovery is off a low base into a structurally lower-margin regime, not back to prior peak economics.
m48
Top line still contracting
Recent revenue YoY -1.7% versus category median recent growth +3.3% — the company is not participating in the modest category upturn on the revenue line, which is the clearest signal that volume/mix in Europe remains impaired.
m39
Macro/demand headwinds on end markets
10y at 4.63 with a flat-ish curve pressures construction starts and auto affordability — MT's two largest demand pools. A European industrial slowdown would strand the fixed-cost base quickly.
m29
Green-steel capex without paid premium
Decarbonization spend is committed while customer willingness to pay a green premium is unproven; that combination compresses returns on the very capex meant to drive growth.
vs expectations: ~6m inline · 1y inline · 2-3y below
The forward growth verdict — is the business itself likely to grow (next 2 quarters / year 1 / years 2–3), judged against its category and against printed expectations. The full horizon ladder + creme renders on the Growth Outlook card above. Not a call on the price (Valuation owns that) or the tape (Sentiment owns that).
Please log in to view trade setups
The Augustus trade-setup read is a members feature.
Log in
Lenses kept deliberately separate — Company Quality (price-agnostic), Valuation (price-conditional), General Sentiment (non-fundamental macro/narrative pressure), AI Impact (structural ~5yr AI exposure), and Growth Outlook (the forward growth verdict). The scores are not blended. Filing-level items (convertibles, lock-ups, customer concentration) are v2 — see each lens's "verify."
Price Prediction
Lower -24.2% v0.6.0 View full prediction →

When we made this prediction on Aug 16, 2026, MT was $73.86. We expect it to be $56.00 by Feb 2027, and we consider it great value under $40.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Aug 16, 2026.

Price when predicted$73.86
Our estimate for Feb 2027$56.00-24.2%
Great value below$40.00
Price history shown (6 Months)

Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.

Community AI Feedback
No community reviews yet for MT. Be the first — hit How to Contribute, have any AI review this page, and paste its take back here.
My Notes personal — only you see this
v1.1.562 · 9b2927c4 · 2026-08-22 16:52:06