Kaiser Aluminum

Phase A — Understand the business

Company Overview

Kaiser Aluminum makes and sells semi-fabricated specialty aluminum mill products — flat-rolled (plate, sheet, coil), extruded (rod, bar, hollows, shapes), drawn (rod, bar, pipe, tube, wire) and certain cast products — into four demanding end markets with high qualification barriers: Aero/HS Products (aerospace/space/defense), Packaging (beverage/food can sheet), GE Products (general engineering / industrial / semis), and Automotive Extrusions.

The economic model is metal price neutrality: KALU passes the underlying aluminum index (Midwest Transaction Price = LME + Midwest Premium) plus most alloy cost through to customers, and earns its profit on the conversion margin — the value added from fabrication. Hence the company reports both Net sales and Conversion Revenue (Net sales − Hedged Cost of Alloyed Metal); Conversion Revenue is the metric that strips out the metal pass-through and shows the real business. This distinction is the single most important thing to understand about KALU — reported revenue swings with the aluminum price even when the underlying business is flat.

• FY2025: Net sales $3,373.0M; Conversion Revenue $1,453.2M; Net income $112.5M; diluted EPS $6.77; Adjusted EBITDA $310.2M.
• Pricing mechanisms: spot (most GE, some Aero), index-based (most Aero + Packaging, ~all Auto), and firm-price (hedged). Firm-price-risk shipments were only 126.7 mmlbs of 1,108 total in 2025 — i.e. ~89% of volume is naturally metal-neutral.
• Customers: blue-chip aero/auto OEMs and tier-ones, can makers, metal service centers. ~70% sold direct to manufacturers/tier-ones, ~30% to service centers. Largest customer = 16% of Net sales in both 2025 and 2024 — a real but managed concentration.
• Footprint: Trentwood (WA) = flat-rolled heat-treat plate/sheet (Aero + GE); Warrick (IN) = packaging coil (one of the world's largest ingot-casting + coating sites; ~75% of packaging shipments are higher-margin coated); ~10 extrusion/drawing plants; IMT subsidiary (Columbia, NJ) = multi-axis CNC / additive for aero-defense. ~3,840 employees, ~65% unionized (USW/IAM/Teamsters).

Supply Chain

Upstream → Kaiser → end customer, named:

• Inputs: primary aluminum + recycled scrap, bought from "a wide array of vendors" (multi-sourced by design), priced off average MWTP = average LME ($1.19/lb in 2025) + Midwest Premium ($0.59/lb in 2025, vs $0.19 in 2024, $0.23 in 2023). The Midwest Premium spike is the tariff story (see Lens 8/10) and it flows straight through to reported Net sales. Energy and specific alloys are also bought under physical-delivery / surcharge arrangements.
• Kaiser conversion: remelt/cast (Kalamazoo MI, London ON, Los Angeles CA, Heath OH cast billet/log; Trentwood + Warrick cast their own rolling ingot) → flat-roll or extrude/draw → heat-treat / coat / slit.
• Downstream named buyers/competitors-as-context: beverage/food can manufacturers (Packaging); aerospace OEMs + tier-ones (Aero); tier-one automotive suppliers (Auto Extrusions); large metal service centers (GE). KALU is "one of the few remaining US-based aluminum semi-fabricated producers that supply the American defense industry," though defense reaches it indirectly through service centers/machine shops, not direct government contracts.
• Chokepoints / single-source risks: Warrick is the single coated-packaging asset — a fourth coating line was just commissioned there; an outage concentrates packaging risk. Trentwood similarly concentrates aerospace plate (Phase VII expansion just placed in service). ~32% of Net sales runs through customer-based supply-chain-financing (receivables sold) — a liquidity lever, not a red flag, but worth noting. Labor is a chokepoint: USW contracts at Richmond VA and Florence AL expire Nov 2026 / Mar 2026 respectively; Newark + Trentwood (83% of USW-SPT workers) run to Sept 2030.

Competitive Advantages (moats)

The moat is qualification + process know-how, not scale. KALU is sub-scale vs its named rivals — Arconic, Constellium N.V., Novelis Inc. (Aero + Packaging), plus Tri-Arrows (Packaging) and Norsk Hydro ASA (GE/Auto), with import competition in heat-treat plate from South Africa, Europe and China. The 10-KForm 10-KA company’s audited annual report to the US regulator. The most complete thing it publishes. is candid: "Some of our competitors are substantially larger, have greater financial resources."

What protects the conversion margin:

• Aerospace heat-treat plate qualification — multi-year supplier qualification, decades of metallurgical R&D, and the migration to monolithic (machined-from-plate) airframe design that favors KALU's Trentwood plate. "Only a few companies" can make aerospace heat-treat plate.
• Coated can-sheet (FDA-compliant) qualification — alloy + coating systems that "take multiple years" to qualify; Warrick's coated capability is the differentiator (75% of packaging shipments coated).
• KaiserSelect® / KPS — branded enhanced-consistency products + a continuous-improvement operating system; switching costs are real for safety-critical aero and food-contact applications.
• Bargaining power is mixed. Against suppliers: weak-to-neutral (commodity inputs, but multi-sourced and pass-through). Against customers: the 16%-concentration customer and the service-center channel cap pricing power; the moat lets KALU hold premium conversion pricing, not dictate it. Net: a durable niche moat on the conversion spread, not a wide moat on the metal.

Segments

FY2025 vs FY2024, by end market — shipments (mmlbs) / Net sales ($M) / Conversion Revenue ($M):

The tell: total shipments fell 5.5% and total Conversion Revenue was dead flat YoY ($1,453.2M vs $1,456.2M), yet Net income jumped 71% to $112.5M. The earnings growth did not come from the operating business — it came from metal price lag (+$93.0M favorable in 2025 vs +$45.0M in 2024) flowing through Adjusted EBITDA. Packaging is the genuine growth engine (coated conversion $/lb rising 0.83→0.97); Aerospace conversion revenue actually shrank in 2025 on destocking before the 2026 recovery.

---

Phase B — Measure performance

Earnings Result (latest print: Q1 2026, quarter ended 2026-03-31)

A genuinely record quarter, and an enormous headline beat — but read the two revenue lines side by side:

• Beat: reported EPS $3.74 vs $1.93 consensus (+94%); revenue $1,106.8M vs ~$1,004.6M est. A beat of that magnitude is itself a flag that the Street does not model the metal-lag/premium swing well.
• What drove it: Net sales +42% is ~three-quarters metal pass-through — Hedged Cost of Alloyed Metal jumped to $702.4M from $414.2M as the Midwest Premium hit records. Conversion Revenue +11% on +6.8% shipments is the real operating improvement, led by Packaging (Conv Rev $157.4M, +24% YoY; coated $/lb 0.98→1.07) and an Aero/HS volume recovery (shipments 61.5 vs 56.3 mmlbs). Favorable Metal Price Lag was ~$36.0M in the quarter alone — a material, non-operating chunk of the EBITDA jump.
• Guidance raised: management now guides Conversion Revenue +10–15% and Adjusted EBITDA +20–30% YoY for FY2026.
• Balance-sheet flags: Net Debt/EBITDA ~2.5–2.8x; FY2025 was FCF-negative (OCF $111.4M < capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. $136.9M = ~−$25.5M, partly rescued by $53.1M of asset-disposal proceeds → reported investing −$77.8M). Inventories rose to $725.2M from $601.9M and trade receivables to $395.2M from $319.7M — partly metal-price inflation of working capital, worth watching as a cash drag if volumes don't follow. Dividend raised to $0.77/quarter (declared 2026-04-13).
• Market reaction: stock rose on the print and has continued to a 52-week / all-time high (~$183–187 late May 2026). The tape is pricing the windfall as durable.

Earnings Calls (sentiment trend)

No transcripts in the research layer (transcripts/ empty) — sentiment is reconstructed from web summaries, labeled ``.

• Trajectory: the tone has inflected hard from defensive (2023–2024: destocking, "capacity adjustment," packaging margin pressure from competitive additions ) to confident (Q4 2025 → Q1 2026: "record," guidance raised, leverage reduction, aerospace recovery + reshoring + packaging mix as the three pillars ).
• Notable: the Q4 2025 beat saw the stock drop — i.e. expectations had run ahead even before Q1 — whereas Q1 2026 the stock rose. Management consistently frames results around Conversion Revenue and Adjusted EBITDA (the metal-neutral metrics) and flags metal price lag as a called-out swing item — to their credit, they do not hide it. The recurring new phrases are "reshoring," "coated packaging," and "operational execution"; the thing they stopped saying is "destocking."

Comps

Peer set: the named rivals from the 10-K. Multiples are ``; where a clean current figure isn't sourced I write n/a rather than fabricate.

Read: on a trailing P/E basis KALU at ~26x trades at a >2x premium to its closest pure-play public peer Constellium (~11x). That premium is partly justified (cleaner balance sheet, higher-quality aero/coated mix, US-domestic-tariff-advantaged) but it is being paid on peak, lag-inflated earnings — the more relevant lens is EV/EBITDA on a normalized (ex-metal-lag) EBITDA, where the multiple looks far richer than the headline 9–10x. Comps do not support chasing the stock here.

Stock-Price Catalysts (last ~5 yrs)

• 52-week range $71.44 → $189.86; ~2.5x off the 2025 low; all-time high $187.03 on 2026-05-27. The dominant move is the 2025→2026 re-rate on the Section-232 / Midwest-Premium windfall + aero recovery.
• The structural catalyst is policy, not the company: Section 232 aluminum tariffs raised to 50% (expanded/increased effective April 6, 2026), creating a domestic "price wall." The US Midwest Premium hit a record ~$2,182/ton in early 2026. Domestic semi-fabricators with US capacity are the relative winners; this is why KALU re-rated.
• Pattern revealed: the market reacts to (1) the aluminum-price/Midwest-Premium regime (macro/policy), (2) aerospace build-rate / destock-restock cycle, and (3) earnings prints vs the (mis-modeled) consensus. Note the Q4 2025 beat-but-stock-fell episode — a sign the name can be a "sell-the-news" at elevated expectations.

---

Phase C — Judge people & books

Management

• Keith A. Harvey — Chairman, President & CEO. Joined Kaiser 1981 as an industrial engineer (Ravenswood WV rolling mill); VP 1994; EVP–Fabricated Products 2014; President & COO Dec 2015; CEO July 2020; Chairman Jan 1, 2025 (succeeding Jack Hockema). A deep operator/lifer, not a financial-engineer parachute. Predecessor Hockema credited him with "the Company's growth and transformation to a position of industry leadership."
• Track record: presided over the Warrick packaging acquisition integration, the Trentwood Phase VII modernization, and the fourth Warrick coating line — capacity/quality investments squarely in the moat. Delivered the destock trough (2023–24) without a dividend cut and is now harvesting the up-cycle.
• Skin in the game / capital allocation: insider ownership only 1.7% ($48M) — modest alignment, typical of a professionalized post-bankruptcy industrial (KALU emerged from the legacy Kaiser asbestos bankruptcy in 2006). Recent insider activity is sell-only: independent director Brett Wilcox sold 56% of his stake ($2.6M) at ~$174 — a notable disposal at the highs. Capital-allocation pattern = disciplined, dividend-first ($51.3M paid 2025, ~flat 3 yrs; raised to $0.77/qtr in 2026), heavy reinvestment ($136.9M capex 2025), no buybacks in 2025 (treasury stock unchanged), and proactive liability management (refinanced the 4.625% 2028 notes into 5.875% 2034 notes). FY25 ROE ~14.4%.
• Red flags: none egregious. Comp tied to Adjusted EBITDA + TSR-vs-peers performance shares (reasonable). The watch-item is the gap between "record results" messaging and the metal-lag/premium composition of those records, plus director selling into strength.
• Archetype: seasoned professional operator running a cyclical at the top of its cycle — competent and shareholder-friendly, but not a capital-allocation wizard with a personal fortune riding on the stock.

Forensic Red Flags

Grounded in filings; figures labeled.

• Accounting-method change (the biggest item): effective Jan 1, 2025 KALU switched inventory valuation from LIFO to WAC, retrospectively restated to all periods (cumulative-effect adjustment to Jan 1, 2023 retained earnings). In a rising-metal environment, LIFO→WAC tends to lift reported margins/earnings vs LIFO — so part of the optical earnings improvement is a methodology change, not operations. Legitimate and disclosed, but it makes multi-year comparisons softer than they look and warrants scrutiny.
• Non-GAAP reliance: the whole story is told in Conversion Revenue and Adjusted EBITDA, both non-GAAP. Adjusted EBITDA includes the favorable Metal Price Lag (+$93.0M in 2025, +$36.0M in Q1 2026) and excludes "non-run-rate" environmental + restructuring costs. The adjustments are transparently reconciled, but a buyer must mentally strip the metal-lag tailwind to see run-rate earnings.
• Cash vs earnings divergence: OCF fell to $111.4M in 2025 from $167.1M (2024) and $211.9M (2023) even as net income rose — driven by working-capital build (inventory $601.9M→$725.2M; receivables $319.7M→$395.2M), much of it metal-price inflation. Earnings up, operating cash down is the classic flag to monitor; here it has a benign explanation (metal-cost inflation of working capital) but becomes a real risk if volumes/premiums reverse and that working capital doesn't unwind cleanly.
• Receivables financing: ~32% of Net sales runs through customer-based supply-chain-finance programs (receivables derecognized as sales) — improves reported DSO/cash; standard but flattering.
• Goodwill/intangibles: small ($18.8M goodwill, $41.0M intangibles) — no impairment risk of note.
• Pension/legacy: multiemployer plans + a ~$4.6M partial-withdrawal liability expected 2027 (Sherman TX exit); manageable.

Regulatory findings (required sub-section).

• SEC Litigation Releases: none. AAERs: none. Verified via SEC EDGAR EFTS (LR + AAER) for 2021-06-18 → 2026-06-18.
• 10-K Item 3 (Legal Proceedings): "None.". (Item 4 Mine Safety — not applicable.)
• Non-SEC enforcement (web search "Kaiser Aluminum" FTC/DOJ/FDA/CFPB/consent-decree/settlement/penalty): no material federal enforcement action surfaced. The only recurring regulatory exposure is environmental remediation of legacy pre-2006 contingencies (handled via undiscounted accruals; "non-run-rate environmental expenses") and evolving climate/packaging-content disclosure regimes — disclosed risk factors, not enforcement actions.
• Verdict: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-18. Clean legal/regulatory record; the only "forensic" caution is the LIFO→WAC switch + metal-lag-flattered non-GAAP, both fully disclosed.

---

Phase D — Project & stress-test

Forward Projection (EPS, FY2026 → FY2028)

Built bottom-up from FY2025 actuals ($6.77 dil EPS, 16.6M dil shares) + the raised FY2026 guidance. No our model create is run in --watchlist mode (per skill). All outputs `` with arithmetic shown.

Anchor facts: FY2025 Adj EBITDA $310.2M (incl. +$93.0M metal-lag); Q1 2026 annualizes hot (~$128.5M × 4 ≈ $514M, but Q1 carried record premium + $36M lag and seasonality favors 2H packaging). Management FY2026 guide: Conv Rev +10–15%, Adj EBITDA +20–30% → FY2026 Adj EBITDA ≈ $372–403M.

Crux: the base case has FY2027 EPS roughly flat-to-down vs FY2026 — because a large slice of 2026 earnings is a non-repeating metal-lag + record-premium windfall. The Street's own consensus PTConsensus price targetThe average of what published analysts think the share price should be. An opinion poll, not a forecast. ($159.50, below the ~$180 spot ) is consistent with this: analysts already model normalization. Brier forecast I'd log if this were a conviction position (not logged in --watchlist): "KALU FY2027 non-GAAP EPS < FY2026 non-GAAP EPS, p≈0.60."

Bull vs Bear

Bull case. A best-in-class US specialty fabricator with genuine qualification moats in aerospace heat-treat plate and FDA coated can-sheet, riding three real tailwinds: (1) Section-232 50% tariffs structurally advantage US-domestic capacity — Bessent has signaled the 50% primary-metal duty is "a permanent fixture"; (2) aerospace up-cycle as build-rates recover and monolithic design favors plate; (3) Packaging coated-mix compounding (the secular plastic→aluminum shift, coated $/lb rising). Clean-ish balance sheet (~2.5x), dividend-first, just-completed growth capex (Warrick line + Trentwood Phase VII) now harvesting. KeyBanc Overweight, PT $183.

Bear case (permanent-impairment-grade risks). (1) The earnings base is partly a windfall — metal-price-lag (+$93M FY25, +$36M Q1) and a record Midwest Premium ($0.59/lb vs $0.19 a year prior) are mean-reverting; a normalization takes EBITDA down even if the business is fine, and the lag can flip negative on falling aluminum. (2) Sub-scale vs Arconic/Constellium/Novelis/Hydro — no pricing power on the metal, perpetual import competition (incl. Chinese plate), and a 16%-concentration customer. (3) Working-capital + FCF fragility — FY25 was FCF-negative pre-disposals; a metal-price reversal could trap inflated inventory/receivables. Pre-mortem (18 months out, thesis broke): aluminum/Midwest-Premium rolled over (tariff carve-outs on finished goods, or demand softening), metal-lag swung negative, aero re-destocked, and the stock de-rated from ~26x peak EPS to ~12x normalized EPS — a double hit (lower E and lower multiple) from a 52-week high. Multiples are too high on normalized earnings. Contrarian view the market is refusing to see: the 2026 "record" is being capitalized as a new baseline, when Conversion Revenue (the real business) grew only ~11% and shipments ~7% — the other ~30% of the EBITDA jump is a policy/price gift that doesn't compound.

Devil's Advocate (short-seller)

Dismantling the bull case. What structurally breaks the model: KALU earns a conversion spread on metal it doesn't control. The entire 2025→2026 re-rate rests on an exogenous, policy-driven aluminum-premium spike + a favorable metal-price-lag — neither is a Kaiser achievement and neither compounds. Strip the +$93M (2025) and +$36M/qtr (2026) lag and the run-rate EBITDA is dramatically lower; the stock is then ~26x trailing EPS that is itself inflated. Revenue concentration: 16% in one customer; ~30% through fickle service centers; aerospace just demonstrated it can destock (Aero Conv Rev −14% in 2025). Weakest-moat angle: in GE and Packaging the company concedes larger rivals and import competition — the moat is narrow (aero plate, coated sheet) and the rest is commoditized conversion. Most dangerous competitor bulls underestimate: Constellium (trading at ~11x, half KALU's multiple, with deeper aero+auto value-add) and imported heat-treat plate if the premium gap ever invites it. Worst capital-allocation / governance tells: insiders own only 1.7% and are selling (a director dumped 56% at ~$174); no buyback to signal the stock is cheap (because it isn't). Accounting: the LIFO→WAC switch coincidentally flatters margins into the up-cycle. What must hold for ~$180: that record premiums + positive lag persist for years AND aero ramps AND multiple stays ~26x. If growth disappoints 20–30% (premium normalizes, aero stalls): normalized EPS ~$5–6, a fair multiple ~12–14x → stock $65–85, i.e. ~50%+ downside — which is exactly where the 52-week low was. Single permanent-impairment scenario: a structural aluminum-price collapse + sustained negative metal-lag during an aero downcycle, trapping inflated working capital — plausible-but-not-base-case (~20–25%).

Management Questions (ordered by information value)

1. Of the ~$385M FY2026E Adjusted EBITDA, how much do you estimate is metal-price-lag + above-mid-cycle Midwest Premium vs. structural conversion-margin — i.e. what is the normalized, premium-neutral run-rate EBITDA?
2. If the Midwest Premium reverted to its 2023–24 range (~$0.20/lb), what is the dollar EBITDA impact, and how much is hedged vs. exposed?
3. What is your explicit assumption for metal-price-lag in FY2027 in the guidance, and what happens to reported EBITDA if aluminum prices fall next year?
4. Why no share repurchases in 2025 — is the dividend-only policy a valuation signal, and at what price would you buy back stock?
5. Aerospace Conversion Revenue fell 14% in 2025 then recovered in Q1 2026 — how much of the Q1 aero strength is true build-rate demand vs. restocking, and what are your contracted 2026–27 aero volumes?
6. Packaging coated conversion $/lb keeps rising (0.83→0.97→1.07) — is that durable mix/pricing or are competitive coated additions about to compress it again?
7. FY2025 operating cash fell to $111M (from $212M in 2023) on working-capital build — when does inflated inventory/receivable cash convert back, and what is your FY2026 FCF target after $120–130M capex?
8. With the Warrick fourth coating line and Trentwood Phase VII now in service, what is the incremental revenue/EBITDA each unlocks at full utilization, and when?
9. What is the realistic single-customer (16%) renewal/loss risk, and how concentrated is aerospace within that?
10. USW contracts at Richmond and Florence expire in 2026 — what wage/benefit inflation is embedded in renewals, and what is the strike-risk contingency?
11. How do you think about the ~26x P/E the market assigns vs. Constellium at ~11x — what does the market understand about Kaiser that it doesn't about peers?
12. Defense reaches you indirectly via service centers — is there an opportunity (or risk) in the current defense-spending mix, and would you pursue direct government qualification?
13. What is your capital-allocation priority order for the next windfall of cash — debt paydown (toward <2x), buyback, dividend growth, or M&A?
14. How exposed are you to imported heat-treat plate (South Africa/Europe/China) if the tariff structure on derivative products is softened to ease consumer inflation?
15. What is the long-term Conversion Revenue growth algorithm (volume vs. price/mix) you're underwriting, independent of metal prices?

---