Umicore

Phase A — Understand the business

Company Overview

Umicore is a Brussels-listed (BEL 20) materials-technology and recycling group — not a miner and not a chemicals major, but a closed-loop metallurgist that sits between the two: it buys metal-bearing feed (ores concentrates, industrial by-products, spent catalysts, e-scrap, end-of-life batteries), extracts and refines ~20 metals, and sells engineered materials back into autos, electronics, and energy.

Post-2024 reorganisation it reports four business groups:

The commercial punchline is in the last two columns: Catalysis + Recycling = ~97% of pre-corporate segment EBITDA (€821M of €908M). Battery Materials — the unit the entire 2020–2023 equity story was built around — is a cash drain the market once paid a growth multiple for. Specialty is a useful, quietly-improving helper. This is a company whose real earnings engine is the two "boring" legs it is often not credited for.

Contract structure — the tell. In Battery Materials, ~11% of FY2025 revenue "growth" came from take-or-pay compensation — customers paying penalties for not lifting contracted CAM volumes, rather than buying product. That is contractual downside protection working as designed, but it is emphatically not demand. Catalysis and Recycling run on pass-through metal economics (Umicore is largely metal-price-neutral on principal, earning a refining/processing margin) with a large precious-metals hedge book — ~70% of 2026–27 metal exposure locked.

Supply Chain

Map, upstream → Umicore → end customer, named where sourced:

Upstream feed (inputs):

• Primary metal / concentrates: legacy mining-concentrate feed largely displaced — Hoboken was re-tooled to run on recyclables and industrial by-products rather than mined concentrate.
• Cobalt: sourced with an OECD-conformant, third-party-assured due-diligence framework — DRC exposure is the latent ESG chokepoint.
• Lithium / nickel / precursors for CAM: externally sourced; Umicore's late, NMC-centric position (vs. the industry's LFP shift) is a strategic weakness (Lens 3).
• E-scrap & spent autocatalysts: the raw material for Recycling — sourced globally from OEMs, refiners, and the electronics waste stream. ~300g gold recovered per tonne of e-scrap (vs. grams per tonne in mining).
• Germanium: feeds Specialty Materials — directly exposed to China's germanium export restrictions (China is the dominant supplier; Umicore is one of the few Western refiners).

Chokepoints / single-source dependencies:

• Hoboken (Antwerp) is a single mega-asset chokepoint for Recycling and PGM refining — one site, irreplaceable in the medium term, with a scheduled major-maintenance shutdown in H2 2026 that will dent Recycling earnings for the period (recurring ~biennial event, framed as manageable).
• IONWAY JV ties Battery Materials to Volkswagen/PowerCo's single-customer battery roadmap — as VW dials battery ambition up and down, Umicore's Nysa (Poland) CAM economics move with it.

Downstream (customers): global automakers and Tier-1s (Catalysis); EV cell makers SK On (contract renewed for 2026), ACC (battery-recycling award), IONWAY/VW (Battery Materials); semis/optics/aerospace (Specialty); the refining/industrial complex (Recycling). Customer file on disk is empty — names above are.

Competitive Advantages (moats)

The genuine moat is Recycling, not batteries. Hoboken is described across trade and academic sources as "the world's most advanced / one-of-a-kind, largest" precious-metals recycling facility, recovering 17 metals from complex feed via lead/copper/nickel collector metallurgy (gold, silver, Pt, Pd, Rh + antimony, bismuth, tin, selenium, tellurium, indium…). This is a process + permit + scale moat that is close to un-replicable: decades of proprietary metallurgy, an operating permit for a hazardous smelter inside a residential zone that no one could newly obtain today, and the flexibility to profitably process feed streams competitors cannot. That is the durable asset.

Catalysis is a strong-but-fading moat. Umicore is a global top-3 autocatalyst maker with deep OEM qualification lock-in (multi-year, safety-critical, hard to switch). The moat is real today but structurally shrinks with internal-combustion volume — a melting-ice-cube franchise with a longer-than-feared runwayRunwayHow long the cash lasts at the current rate of spending. It shortens the moment spending rises, which is why a figure taken from a quiet quarter flatters. (see Lens 5). Bargaining power is balanced: OEMs are concentrated, but re-qualifying a catalyst supplier is slow and costly.

Battery Materials has, at best, a contested moat. The Western-CAM-champion thesis assumed cost/IP/scale advantages that Chinese cathode makers erased on cost and that the industry's pivot to LFP (where Umicore was late) undercut on chemistry. The €1.6B impairment (Lens 5) is the market's verdict on that moat. What remains is IP, a Western/localised supply-chain option (IRA/EU local-content tailwinds), and take-or-pay contract protection — call options, not a moat.

Bargaining power, net: strong over e-scrap suppliers and in PGM refining (few alternatives to Hoboken); weak in CAM (buyers hold the cost whip, chemistry is commoditising).

Segments

FY2025 segment detail (ex-metal revenue / adj. EBITDA / adj. EBIT), all:

Segment EBITDA sums to €908M; group adj. EBITDA is €847M — the ~€61M gap is Corporate/eliminations. Trend: the mix is de-risking — the two structural earners are stable-to-growing, and the battery drag is shrinking. But the quality of the improvement is cyclical (metals) and contractual (penalties), not secular growth.

---

Phase B — Measure performance

Earnings Result (FY2025, reported 2026-02-20)

The full multi-year arc — this is a peak → impairment → trough → recovery story:

All rows.

FY2025 hard numbers:

• Group revenue (ex-metal) €3.6B (+3%); turnover incl. metal €19.4B
• Adj. EBITDA €847M (+11%), margin 24.0% (from 22.0%); adj. EBIT €579M (+21%)
• Adj. net profit (group share) €288M; adj. EPS €1.20 (+13% YoY per the call)
• Cash flow from operations €849M; free operating cash flow €524M → ~11% FCF yield on ~€4.75B cap
• Capex €310M — the headline of the whole turnaround (vs. €800M–1B+ at the battery-build peak; guidance was €400M, so they under-spent)
• Net debt €1.4B; net debt/EBITDA 1.60x (down from ~1.9x); ROCE 15.7%
• Dividend €0.50 reset as the new baseline (interim dividend discontinued), ~42% payout, 28-year uninterrupted dividend streak

Drivers: the beat was cost + capital discipline (€100M efficiency delivered) and metal-price tailwind, not top-line growth. Balance-sheet flag worth noting: Umicore monetised its permanent gold inventory for ~€525M cash and switched to metal leasing — a real de-leveraging lever, but one that trades a balance-sheet asset for a recurring lease cost.

2026 outlook: management guided FY2026 adj. EBITDA to "approach €1 billion" at the Q1 update (2026-04-30), assuming stable metal prices and no macro shock — a ~+18% step-up vs. €847M. Note the gap: sell-side consensus for FY2026 group adj. EBITDA was €880.7M as of 2026-04-29 — i.e. the company's own signal sits ~13% above the pre-Q1 consensus. Either consensus catches up, or the €1B is aspirational.

Market reaction pattern: the 2024 impairment/guide-down was the biggest BEL 20 faller of its day; FY2024 results took the stock down >10%. Through 2025–26 the stock recovered ~55% off the €12.77 low as the reset landed and metals rallied.

Earnings Calls (sentiment trend)

Tone has traveled crisis (2024) → repositioning (2025) → cautious confidence (2026):

• What management now emphasises: "cash discipline," "solidifying the balance sheet structurally," "efficiency as DNA" (€100M delivered), "selective/phased capital," and contract enforcement over volume-chasing in batteries. Target leverage stated as 1.5–2.0x.
• What they stopped saying: the 2021–2023 language of "battery growth champion," aggressive CAM capacity build-out, and hockey-stick EV volume ramps. The battery narrative is now value recovery, not growth.
• New/recurring phrases: "take-or-pay compensation" (now material), "hedging visibility," and — notably — AI-driven demand for data-centre backup-power catalysts and fuel-cell (PEM) catalysts as fresh Catalysis options.
• Tells: management declined to give numerical guidance at FY2025 ("market still very dynamic… we would expect adjusted EBITDA to further progress"), then warmed to the "approach €1B" language only after a strong Q1 — a confidence-building sequence, but also a reminder that visibility is thin. On IONWAY, they deflected VW-scaling-back concerns with "clear contracts in place" — a defensive posture on the single biggest battery dependency.

Comps

Umicore has no clean single peer — it is a three-way mongrel (autocatalysts ≈ Johnson Matthey; precious-metals recycling ≈ Aurubis/Boliden; CAM ≈ Korean cathode makers/BASF). Multiples below are with date or n/a.

Read: Umicore screens cheap on cash flow (~11% FCF yield, ~6–7x forward EV/EBITDA) and mid-teens on adjusted P/E. The instructive comp is Johnson Matthey: the closest pure autocatalyst peer is loss-making and dismantling itself, with a strategic buyer (Honeywell) paying £1.8B to consolidate the process-catalyst piece. That validates two things — (1) auto-catalyst cash flows are real and coveted, and (2) the market treats these franchises as melting assets to be harvested/consolidated, not growth stocks. Umicore's premium to JM is earned by Recycling (the crown jewel JM lacks) and a still-positive P&L.

Stock-Price Catalysts (5-yr, moves >5%)

What actually moves this stock: (1) the battery/EV narrative and impairment risk (the violent moves are all battery-driven), and (2) precious-metal price cycles (the quieter grind). Quarterly beats/misses matter far less than those two macro dials. Today the tape is trading a balance-sheet-repair + metals-cycle re-rating — which is precisely why the durability of metal prices is the whole ballgame.

---

Phase C — Judge people & books

Management

CEO — Bart Sap (since 2024-05-16). A 20-year Umicore insider: joined 2004 as controller for Cobalt & Specialty Materials; roles across Korea and Belgium in finance, supply, and refining of battery raw materials, business development and operations; SVP Cobalt & Specialty Materials & Supply (2020); EVP Catalysis (Mar 2021) — i.e. he ran the profitable core, not the battery overbuild. Master's in Commercial Science (Accountancy & Taxation), Vlekho Brussels.

• Archetype: finance/operations "safe hands," not a visionary. Exactly the profile a board installs to stabilise, cut capex, and harvest cash after a growth bet fails — and that is precisely what FY2025 delivered (capex €310M, €100M efficiency, leverage 1.9x→1.6x, dividend reset and defended). Track record so far: credible on discipline, unproven on re-igniting growth.
• Predecessor — Mathias Miedreich (ex-Faurecia, external): the battery-growth champion who oversaw the CAM capacity build and the €1.6B impairment's underlying decisions; stepped down "in mutual agreement" May 2024 — effectively the board changing horses after the strategy blew up.
• Chairman (Supervisory Board): Thomas Leysen — long-standing Belgian corporate figure. CFO: Wannes Peferoen (per the FY2025 call).

Capital allocation — the reset is the story. From ~€2.1B of planned 2025–28 capex being cut (incl. ~€1.4B off battery), Flanders battery ops closed, the Canada CAM plant halted, and ~260 jobs removed. That is textbook value-preservation after over-investment. Insider ownership / insider-transaction data: n/a (no our figures on disk).

Red flags (management): (1) the prior regime's capital destruction is fresh — €1.6B written off on a strategy the same board approved; (2) the IONWAY commitment (€500M through 2026) keeps Umicore chained to VW's wavering battery plans even mid-retrench; (3) selling the permanent gold inventory (€525M) flatters the balance sheet but is a one-time lever. None are governance-integrity red flags — this is a strategy-error story, not a fraud story.

Forensic Red Flags

Accounting-risk read (web-only — no filings on the shelf; figures ):

• Impairment / goodwill-intangibles: the €1.6B FY2024 battery impairment (PPE + non-current inventories, mainly Asia) is the dominant item — already taken, which de-risks future surprises but signals prior over-capitalisation. Watch for further battery write-downs if the EV recovery slips again (management already guides battery EBIT below break-even through 2026).
• Revenue quality: the flag is take-or-pay compensation driving ~11% of Battery Materials revenue growth — legitimate under IFRS but low-quality, non-recurring-flavoured revenue dressed as recovery. Scrutinise how much of any "battery turnaround" is penalties vs. volume.
• Metal / inventory & hedging: Umicore runs a large hedge book (~70% of 2026–27 exposure locked) and monetised permanent metal inventory (€525M) into leasing. This stabilises earnings but caps upside and introduces lease-cost and mark-to-market complexity; cash-flow vs. earnings can diverge on working-capital swings from higher metal prices (management flagged working-capital pressure).
• Cash vs. earnings: FY2025 CFO €849M vs. adj. EBITDA €847M and FOCF €524M — cash conversion is strong and clean, a genuine positive that argues against aggressive accounting.
• SBC / non-GAAP: adjusted vs. statutory gap is driven by impairments/restructuring, not by SBC inflation — normal for a company mid-restructuring. Detail n/a (no filings on disk).

Regulatory findings (required):

• SEC (EDGAR): none — Umicore has no CIK and does not file with the SEC; no EDGAR/LR/AAER search is possible.
• Non-SEC enforcement — one material antitrust finding: the French Autorité de la concurrence fined Umicore ~€69.2M for abuse of a dominant position in rolled-zinc building products (VMZINC) — an exclusivity policy toward distributor-resellers running ~1999–2007 (and per the authority persisting to 2014). A real competition finding, though in a peripheral (building-products) line, not the core metals franchises.
• Environmental / social license — Hoboken lead: the Hoboken smelter has a decades-long lead-emissions overhang — children's blood-lead monitored since 1978, a 2020 spike, ~€25M/yr environmental spend, and voluntary house buy-backs to create a "green buffer." Recent readings improved (~2.70 µg/dL). No lawsuit or financial settlement was sourced, but this is a live operating-permit and reputational risk for the single most important asset in the group.
• Cobalt / conflict-minerals: OECD-conformant, third-party-assured due-diligence framework — a mitigant, not a penalty; latent DRC-sourcing ESG risk remains.
• Net: no accounting-fraud or securities findings; one €69M antitrust fine in a non-core line; a genuine environmental/social-license overhang at Hoboken. Verified via the research-layer regulatory file (SEC: none) + web search (French competition authority, Hoboken) as of 2026-07-10.

---

Phase D — Project & stress-test

Forward Projection (EPS, next 3 fiscal years)

Anchored on FY2025 actuals (adj. EBITDA €847M, adj. EPS €1.20, ~240M shares) and the FY2026 "approach €1B EBITDA" signal. All outputs; provenance on inputs inline. Per --watchlist rules, no our model forecast is logged.

Sell-side reference points: FY2026 group adj. EBITDA consensus €880.7M (MarketScreener, 2026-04-29); FY2026 adj. EPS consensus ~€1.65 (Bloomberg) vs. JPM €1.82 (OW, PT €27) vs. a basic-EPS broker figure of ~€1.30 — the spread itself flags low visibility. My base ~€1.55 sits deliberately below the bullish house numbers and roughly at/under consensus: I do not fully credit the €1B, because a meaningful slice of it is cyclical-metal and take-or-pay in quality.

The forecast that matters is not EPS precision but the durability of the two harvest legs: if gold/PGM hold and ICE grinds down slowly, base is right; if either cracks, bear is fast.

Bull vs Bear

Bull case (narrative). Umicore is a balance-sheet-repair-plus-metals-cycle re-rating that the market is still pricing as a broken battery stock. The €1.6B impairment cleared the decks; a disciplined insider CEO cut capex by two-thirds and reset the dividend to a defensible €0.50 baseline; leverage fell to 1.6x; FCF yield is ~11%. On the two legs that actually earn — a near-irreplaceable Hoboken recycling moat riding record precious-metal prices, and an autocatalyst cash cow enjoying a longer-than-feared ICE/hybrid runway — 2026 EBITDA is heading toward €1B. Battery Materials is a free call option now managed for cash (take-or-pay floors, IONWAY within budget) with IRA/EU local-content optionality. At ~6–7x forward EV/EBITDA and ~12x forward earnings with a covered ~2.5% yield, the downside looks priced and the re-rating has room to ~€24–27.

Bear case (2–3 permanent-impairment risks). (1) The €1B is peak-quality, not run-rate — Recycling is levered to record gold/PGM (hedged ~70%, so capped on the upside and exposed to mean reversion), and Catalysis is a structurally declining ICE franchise on borrowed time; strip cyclical metals and ICE overtime and the normalized earnings base is lower. (2) Battery Materials could need more write-downs — if the EV recovery slips again or LFP keeps taking share from NMC, the "recovery to above-cost-of-capital by decade-end" quietly slips further, and IONWAY becomes a cash sink chained to VW's wavering plans. (3) Hoboken single-asset + permit/social-license risk — one environmental escalation at the crown-jewel site is an outsized, un-diversifiable tail.

Pre-mortem (18 months out, thesis broke): gold/PGM rolled over 15–20%, taking Recycling EBITDA down with them; the H2-26 Hoboken shutdown ran long; European ICE production fell faster than modeled, hitting Catalysis volumes; battery losses widened rather than narrowed; consensus that had crept toward €1B re-based to ~€820M, and the stock round-tripped to the low-teens.

Multiples too high? No — the multiple is undemanding. The risk is the denominator (earnings quality/durability), not the multiple.

Contrarian view (what the market refuses to see): the market frames Umicore as a transformation (batteries recovering). The truer frame is that Umicore is worth more with the battery dream dead than alive — it's a cash-returning metals-recycling-and-catalyst harvester, and the 2026 re-rating is powered by the "boring" legs, not the story the bulls tell. Own it for the metals cycle and capital return, and demand a discount for the fact that both harvest legs are near cyclical peaks.

Devil's Advocate (short-seller)

Dismantling the bull case:

• What structurally breaks the earnings? Two of the three earning legs are cyclical or secularly declining. Recycling's ~€371M EBITDA is a function of record precious-metal prices — the hedge book (70% locked) proves management itself doesn't trust the price to hold. Catalysis's €450M is auto-catalyst demand on an ICE base that must shrink — the only debate is speed. That's ~97% of pre-corporate EBITDA sitting on a metals-cycle peak and a melting franchise.
• Concentration: Battery Materials is dangerously single-customer-cluster dependent (SK On, IONWAY/VW). VW is retrenching on batteries; "clear contracts in place" is exactly what you say when your counterparty is wobbling.
• Weakest-moat claim: the CAM "moat" was already disproven — €1.6B says so. Chinese cathode cost leadership and the LFP shift are the dangerous competitors bulls keep hand-waving; Umicore is a sub-scale, late, high-cost Western NMC player in a chemistry the market is leaving.
• Worst capital-allocation moves: the entire 2021–2023 battery build now written down; €500M still going into IONWAY while retrenching; monetising the permanent gold inventory (€525M) — a one-off that makes the deleveraging look better than the underlying trajectory.
• What must hold for the price? That gold/PGM stay near records and ICE declines slowly and batteries stop bleeding. Three cyclical/secular bets, all pointing the same way.
• Earnings disappoint by 20–30%? On the bear path (~€760–820M EBITDA, EPS ~€1.00–1.10), a de-rating to ~5x EV/EBITDA puts the stock back to the low-teens — i.e. the 52-week low is a real gravitational center, not a floor left behind.
• Single permanent-impairment scenario: a serious Hoboken environmental/permit event — the recycling moat is the un-permittable single site. Plausibility: low-probability, high-severity; the reason a discount is warranted.

Management Questions (ordered by information value)

1. Of the FY2026 "approach €1B" adj. EBITDA, how much is structural run-rate vs. cyclical precious-metal price and non-recurring take-or-pay compensation — what is the normalized number at mid-cycle metal prices?
2. What gold/PGM price is embedded in the 2026 plan, and what is Recycling EBITDA sensitivity to a 15% decline given the ~70% hedge?
3. On the batteries: what is the honest year you expect Battery Materials to earn above its cost of capital — and what EV-volume and LFP/NMC-mix assumptions does that require?
4. If VW/PowerCo further scales back, what are Umicore's contractual protections and exit economics on IONWAY, and how much more capital is committed beyond the €500M?
5. What is the normalized decline rate you model for auto-catalyst volumes as ICE production falls, and when does Catalysis EBITDA peak?
6. Is there any residual battery-impairment risk on the current asset base under a slower-recovery scenario?
7. How large and durable are the fuel-cell and data-centre backup-power catalyst opportunities — real EBITDA lines this decade, or optionality?
8. What is the through-cycle capex the group needs, now that the battery build is cut — and what does that imply for sustainable FCF and buyback capacity?
9. What is the replacement/insurance and permit-continuity plan for Hoboken given single-site concentration and the environmental overhang?
10. After monetising the permanent gold inventory, what is the recurring lease cost, and are there more balance-sheet levers of that kind left?
11. What would make you raise the dividend above €0.50, and how do you weigh dividend growth vs. buybacks at ~11% FCF yield?
12. Where can the €100M efficiency program go from here — is there a second wave, and at what cost to capability?
13. How exposed is germanium/Specialty Materials to China export controls, and does that cut both ways (supply risk vs. Western-refiner pricing power)?
14. What is your M&A appetite — are you a consolidator of recycling/catalyst assets (à la Honeywell/JM), or purely deleveraging?
15. What is the worst-case cash cost of the Hoboken lead/environmental remediation and social-license program over the next decade?

---