Cleveland-Cliffs Q3 2026 Earnings: $165M Profit, $612M EBITDA and a $4B 2027 Guide

Cleveland-Cliffs reported its strongest quarter in three years on October 19, 2026: third-quarter revenue of $5.6 billion (+18.5%), GAAP net income of $165 million ($0.26 a share) against a year-ago loss, and adjusted EBITDA of $612 million — above the $575 million it had guided. The steelmaker shipped 4.2 million net tons at an average price of $1,177 a ton, then told investors 2027 adjusted EBITDA should exceed $4 billion. After two straight annual losses totaling $2.2 billion, the cycle has turned. The question is how much of it is price, and how much is Cliffs.

The Headline Numbers​

Cliffs' fiscal year is the calendar year, so Q3 2026 is the three months ended September 30, 2026. Every figure below comes from the Form 10-Q and the earnings release, whose condensed statements agree with the 10-Q face statements line for line. Net income throughout this post is attributable to Cliffs shareholders ($152 million; consolidated net income was $165 million before the $13 million noncontrolling-interest share).

Read the table from the bottom up and the quarter explains itself. Volumes grew 3%; prices grew 14%; revenue grew 18.5%. That price-over-volume mix is the whole story: $878 million of new revenue against a cost of goods sold that rose only $367 million, flipping gross profit by $511 million. Below the line, the quarter is strikingly clean — no asset-impairment charge, no miscellaneous loss, restructuring of just $1 million. A year ago the same lines carried $25 million of miscellaneous losses; nine months ago they carried a $39 million impairment. This quarter, the P&L is just steel sold at higher prices.

The sequential turn is just as sharp. Against Q2 2026, revenue rose 7.4% ($5,226M to $5,612M), net income swung from a $134 million loss to a $165 million profit, and adjusted EBITDA more than doubled from $286 million to $612 million — the $326 million improvement the release headlines. CEO Lourenco Goncalves' summary — "Average selling price, sales volumes, and unit costs all moved in the right direction relative to Q2 performance" — is one of those sentences you can verify in about thirty seconds, and it checks out on all three.

Revenue Deep Dive​

Distributors did the heavy lifting: up 42%, contributing $541 million of the $878 million revenue gain — nearly two-thirds of it. That is the spot-market channel, the one most exposed to index prices, and it confirms what the shipment data says. Hot-rolled shipments surged 28% to 1,899 thousand tons while the company-wide average price rose 14%: when distributors restock into a rising price, Cliffs' top line moves fast. Infrastructure (+16.5%) and automotive (+13.6%) grew solidly behind it, and the release notes direct automotive shipments were the highest since Q3 2023 — a data point that matters because automotive is the high-margin, contract-protected book (40–45% of flat-rolled shipments sell under fixed-price contracts).

The one declining line is the most interesting. Sales to steel producers fell 10%, and slab shipments — the semi-finished product Cliffs sold ArcelorMittal USA under a legacy contract — went to zero in the quarter from 375 thousand tons a year ago. That contract, roughly 10% of sales volume and explicitly "unprofitable in 2025," concluded in December 2025. Cliffs deliberately walked away from its worst tons. When a company growing revenue 18.5% is simultaneously shrinking its lowest-margin line to nothing, the mix math works twice: fewer bad tons, more capacity for good ones.

Management's market commentary is unusually quotable this quarter, and the ledger backs it up. "Imports remain well below historical levels, lead times are extended, and our customers increasingly recognize the value of a reliable domestic supply chain," Goncalves said — textbook tight-supply language. The 10-Q quantifies it: finished steel imports "significantly below historical levels," year-to-date imports down 15%, and domestic HRC averaging $1,195 a ton, up 42% and the highest quarterly average since Q2 2022. On demand, the release is more restrained — "increasing demand" rather than any sold-out rhetoric — but receivables tell the pricing story anyway: accounts receivable jumped 50% in nine months (more on that in the ledger section), which is what happens when you sell more tons at higher prices faster than customers pay.

The Margin Story​

Two years of negative margins, reversed in one quarter. The mechanics are pure operating leverage on price: gross margin swung nearly 13 points from full-year 2025 (−4.6%) to this quarter (+8.3%) because the $145-a-ton price increase fell almost entirely to the bottom line. Cost of goods sold per ton actually fell slightly even as volumes rose — the payoff of what the 10-Q describes as continuous year-over-year cost-per-ton reductions since 2023, from footprint optimization, lower coal and alloy costs, and working through higher-cost inventory. Price up, unit cost down: that is the entire margin story in six words.

The pricing outlook language is the forward half of the same trade. The company guides Q4 adjusted EBITDA of approximately $715 million, "an increase from the third quarter as a result of increased pricing, partially offset by seasonality, increased maintenance outages, input costs and incentive compensation accruals." Note the ordering: pricing first, and everything else a partial offset. With 40–45% of shipments on fixed-price contracts repricing into an HRC tape 42% above last year, contracted tons provide a delivery mechanism for spot strength to keep flowing into the P&L for quarters. The margin expansion has a pipeline, not just a print.

The One Big Question: Can 2027 Really Print $4 Billion of EBITDA?​

"Based on the new contracts we have already signed with our clients and prevailing index price forecasts, we would expect our 2027 adjusted EBITDA to exceed $4 billion." That sentence is doing more work than any other in the release, so let's size it. Q3's $612 million annualizes to about $2.4 billion; Q4 is guided to $715 million. Hitting $4 billion means averaging roughly $1 billion a quarter next year — 63% above the Q3 run rate. It is a big ask, but the bridge components are all named: signed 2027 contracts (volume and price visibility), prevailing index forecasts (HRC holding near $1,195), the continued mix shift away from slab, and a full year of the optimized footprint.

Start with how far the gap between GAAP and adjusted has already closed — a useful honesty check on the number being guided:

$152 million of GAAP becomes $161 million adjusted — a 6% haircut-to-premium, not a reinvention. Adjusted EBITDA of $612 million reconciles from $618 million of EBITDA the same clean way. When a management team guides an adjusted number, you want the adjustments small and boring. These are.

The second half of the question is the balance sheet, because the $4 billion promise comes attached to a leverage target: under 2.5x net debt to EBITDA by the middle of next year, via "amplified free cash flow in Q4 and property sale proceeds" driving "a major debt reduction in the fourth quarter." Net debt at quarter-end was about $7.6 billion ($7,645M of long-term debt against $77M of cash) — roughly 7.8x trailing-twelve-month adjusted EBITDA of just under $1 billion, but under 2x the guided 2027 figure. The target is credible if and only if the EBITDA materializes; at the Q3 run rate, Cliffs would need to repay roughly $1.5 billion to get there. Watch Q4: the $715 million EBITDA guide plus asset-sale proceeds is the quarter where the promise starts converting into cash. A side door exists — the POSCO memorandum of understanding, which the 10-Q says could bring "substantial proceeds that could be used to reduce our outstanding indebtedness" — but it is unsigned, unpriced, and years from certain.

Tracking a $5.6B Steelmaker in Plain Text​

A steelmaker is an ideal plain-text subject: tons in, dollars out, and every cost visible. Modeling Cliffs in Beancount forces each quarter's story to reconcile — revenue, all nine cost lines, tax, discontinued operations and the noncontrolling-interest share must sum to exactly zero, or the ledger refuses to validate. Here is the third quarter as the ledger records it (income negative, expenses positive, per Beancount convention):

; FY2026Q3 Income Statement — three months ended September 30, 2026
; Check: −5,612 + 5,147 + 153 + 1 + 148 − 64 + 15 − 1 + 46 + 2 + 13 + 152 = 0 ✓
 
2026-09-30 * "Cleveland-Cliffs Inc." "FY2026Q3 Income Statement"
  Income:Revenue                           -5,612 MUSD  ; Revenues: steelmaking 5,447 + other businesses 165
  Expenses:CostOfRevenue                    5,147 MUSD  ; cost of goods sold
  Expenses:SellingGeneralAdministrative        153 MUSD  ; selling, general and administrative expenses
  Expenses:ImpairmentsAndRestructuring            1 MUSD  ; restructuring and other charges
  Expenses:Interest                             148 MUSD  ; interest expense, net
  Income:OtherNet                               -64 MUSD  ; net periodic benefit credits other than service cost component
  Expenses:OtherNet                              15 MUSD  ; changes in fair value of derivatives, net (MinnTac option)
  Income:OtherNet                                -1 MUSD  ; other non-operating income
  Expenses:IncomeTax                             46 MUSD  ; income tax expense
  Expenses:DiscontinuedOperations                 2 MUSD  ; loss from discontinued operations, net of tax
  Expenses:OtherNet                              13 MUSD  ; net income attributable to noncontrolling interests, deducted as the filing does
  Equity:Adjustments                            152 MUSD  ; net income attributable to Cliffs shareholders offset (RE set by balance assertion)

Two postings deserve attention. The $64 million pension-credit line — larger than SG&A growth, larger than the tax bill — is a recurring feature of this P&L, a legacy pension surplus flowing through earnings every quarter; it flattered 2025's losses on the way down and flatters 2026's profits on the way up. And the $13 million noncontrolling-interest deduction is the wedge between the $165 million consolidated profit the release headlines and the $152 million attributable profit shareholders actually own — small today, but the ledger deducts it explicitly every period so the two can never be confused. Our full modeling conventions are in how we model every company.

The balance-sheet number that tells the quarter's story is receivables: $1,442M at year-end to $2,159M at September 30 — up 50% in nine months. The 10-Q attributes it to "rising steel prices and the timing of sales in the period," which is the pricing story capitalized on the balance sheet: Cliffs shipped more tons at higher prices faster than the cash came back. Meanwhile inventories actually fell $78M despite the revenue surge (higher-cost stock working through), and long-term debt rose about $390M on ABL draws to fund the working-capital build — before Q3's $61M of net ABL repayments started the guided paydown. Price up, receivables up, debt to be paid down from the cash conversion: the ledger shows the cycle mid-turn.

The Multi-Year Arc​

The arc is a textbook steel cycle with a leveraged acquisition drawn over it. FY2023 earned $385 million on $22 billion of revenue; then prices collapsed, Stelco arrived in November 2024 (debt more than doubling to $7.1 billion), and 2025 bottomed with an $860 million gross loss — Cliffs selling steel below cost — plus a dilutive October rescue: 75 million shares at $12.69 to repay the ABL. Adjusted EBITDA fell from $1.9 billion to $37 million in two years. The first nine months of 2026 have already printed $993 million of it back, with operating income positive again. The ledger opens at FY2024 — two audited annuals plus the current quarter, the window spanning the Stelco deal, the trough, and the turn. (FY2023 comparatives above come from the FY2025 10-K; FY2024 keeps its own 10-K's as-filed figures, which the FY2025 10-K later restated by $6 million of net loss — disclosed in the ledger banner.)

Debt is the line that has not turned yet: $7.6 billion, its highest print in this window, against equity that shrank to $6.0 billion. But the direction of travel changed in Q3 — net ABL repayments, $224 million of quarterly operating cash flow against $143 million used a year ago, and $3.2 billion of liquidity with no senior-note maturities before 2029. The leverage that amplified the downturn is now the leverage amplifying the recovery.

The Verdict: Bull vs. Bear​

Bull Case

• The price tape is doing the work, and it is contracted. HRC averaged $1,195/nt (+42% YoY, highest since Q2 2022) with 40–45% of shipments on fixed-price contracts — Q4's $715M guide and 2027's $4B+ are substantially pre-sold, not hoped for.
• Mix is improving structurally, not just cyclically. The unprofitable ArcelorMittal slab book (10% of volume) is gone to zero tons; distributors (+42%) and automotive (best shipments since Q3 2023) are the growing channels.
• Costs are falling while prices rise. Year-over-year cost-per-ton reductions since 2023, six 2025 idlings/closures plus the Hamilton finishing idle concentrating tons on profitable HRC — operating leverage cuts both ways and is now cutting favorably.
• The balance sheet has runway and a plan. $3.2B of liquidity, no maturities before 2029, Q3 ABL repayments begun, and a deleveraging target (under 2.5x by mid-2027) that the guided EBITDA makes arithmetic, not aspiration.
• Policy is a tailwind with legislation behind it. 50% Section 232 steel tariffs, expanded to derivatives in April 2026; imports down 15% year-to-date; extended lead times confirm domestic demand has nowhere else to go.

Bear Case

• $4 billion needs a perfect year. It requires ~$1B/quarter against a $612M Q3 print — 63% growth — with "prevailing index price forecasts" doing heavy lifting. Steel forecasts have embarrassed stronger hands than this.
• The "extended lead times" and "increasing demand" language is softer than a true sold-out market. No "demand exceeds supply" claim appears anywhere in the release; direct automotive depends on North American light-vehicle builds that actually fell year over year (3.8M vs 4.0M units).
• Stelco's turnaround is half done. Yes, its strongest quarter since acquisition — but the galvanizing lines are being idled indefinitely because Canada's coated market is "clearly oversupplied," with no government action expected. Half the Canadian bet is still broken.
• $7.6 billion of debt at ~$600M of annual interest needs the EBITDA to arrive on schedule. Trailing leverage is still ~7.8x; a one-quarter price relapse turns the deleveraging story back into a covenant-and-liquidity story.
• The share count is permanently higher. The October 2025 offering added 75M shares (+14%); every per-share figure from the last cycle peak must now clear a bigger denominator, and the 2027 guide is conspicuously not per-share.

Our Take: Buy the turn, underwrite the contracts — but grade the $4 billion as guidance, not earnings. Q3 is the real thing: broad-based, mix-improving, cost-supported, and clean below the line. The 2027 guide is plausible precisely because it is contractual rather than cyclical — signed customer contracts plus a tariff-walled domestic market is the closest a commodity producer gets to recurring revenue. The ledger will tell us within two quarters whether receivables convert to the promised debt paydown; if Q4 prints $715 million and the ABL balance falls by hundreds of millions, the leverage that terrified 2025 becomes the compounding engine of 2027. Until then, this is a show-me story at an attractive starting price — not a victory lap.