Century Aluminum (CENX) Q3 2026 Earnings Call Transcript

Century Aluminum reported Q3 net sales of $961.2 million and adjusted EBITDA of $331.6 million as shipments rose 33% sequentially to 173,483 tonnes following production restarts and expansions; the company ended the quarter with $755 million in cash against $486 million in debt. Management forecast Q4 adjusted EBITDA of $275 million to $300 million, including a $35 million to $45 million sequential pricing headwind and a $5 million to $10 million raw-material headwind. The Oklahoma smelter project advanced toward a final investment decision, with financing terms largely agreed and a plan to reduce permitted fluoride emissions by 37%; management said shareholder returns would be considered after FID.

Analysis

The important change is operating leverage, not the quarter’s headline EBITDA: stable production at restarted assets can improve fixed-cost absorption into 2027. But cash generation is less repeatable than the headline suggests—tax-credit receipts, inventory release and insurance recoveries are timing benefits, while hedge settlements are not a durable operating source. The Q4 guide also makes near-term earnings sensitivity to lagged LME and regional premiums explicit; strong spot prices alone may not prevent a sequential reset.

Over the next 1–3 months, watch whether Iceland’s transformer commissioning and Mt. Holly’s ramp translate into saleable volume and normalized mix, and whether Jamalco’s bauxite plan improves. February guidance should clarify 2027 production, premiums and cash uses. Over 6–18 months, Oklahoma is a material option on U.S. supply tightness, but also a capital-allocation overhang: FID depends on completing power, financing, import-incentive and permitting work, and shareholder returns are explicitly subordinate. Management’s confidence is not confirmation of approvals or final economics.

Contrarian point: the market may be discounting a peak-cycle quarter and project risk, while underweighting the persistence of improved plant utilization. Conversely, net cash should not be capitalized as wholly distributable given the planned project. The article’s cash figures conflict ($713 million in the takeaways versus $755 million in the transcript); reconcile to filings before modeling liquidity. CENX’s reported interest in TeraWulf’s data-center project is contingent optionality, not a reason to treat WULF as a direct aluminum read-through.

Key Decisions for Investors

• CENX: Avoid chasing a post-results spike; consider scaling into a small long on weakness only if LME and regional premiums remain supportive. Reassess after February guidance. The thesis weakens if realized pricing drives EBITDA below the guided range or production ramps fail to deliver saleable volume.
• Model normalized cash flow separately from reported cash conversion: exclude working-capital release, insurance proceeds and episodic tax-credit receipts from the run-rate, and verify the conflicting cash balance against the filing.
• Treat Oklahoma as a catalyst, not base-case value: track FID, executed power and financing terms, import-incentive mechanics and regulatory progress. A delay, materially higher equity requirement or adverse tariff change would undermine the upside and could prolong the capital-return overhang.
• Do not trade WULF as an aluminum proxy. Its project connection is contingent on data-center completion and monetization; verify operating milestones and the terms of CENX’s interest before assigning value.