Albemarle

Refresh delta (since 2026-06-17)

What changed

1. Q2 2026 was the strongest quarter since the 2023 bust. Net sales were $1,743M (+31%), GAAP net income $480M (Q2-25: $23M), GAAP diluted EPS $3.52, adjusted EPS $3.75 and adjusted EBITDA $858M (+155%, a 49% margin) [8-K Q2]. Energy Storage sold 65kt LCE (+11%) at $19.53/kg realized (+60.5%), and its adjusted EBITDA rose 229% to $723M. Specialties grew sales 20% and adjusted EBITDA 61% [8-K Q2]. First-half adjusted EBITDA was $1.52B, adjusted EPS $6.70 [8-K Q2].
2. Outlook: better where Albemarle controls it, weaker in Q3. Specialties guidance was raised to $1.4–1.6B of sales and $275–325M of EBITDA. CapexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. was cut again, to ~$500M from $550–600M. The price-scenario table was left unchanged ($2.4–2.6B of EBITDA at ~$20/kg) [8-K Q2]. On the call the CFO said results were tracking the high end of the $20 case, but that Q3 sales and EBITDA will fall sequentially on lower volumes and an assumed lower price. The high end of the $20 case implies H2 EBITDA of ~$1.08B, about 30% below H1.
3. A fire at Greenbushes. On 2026-06-09 a fire hit Talison's third chemical-grade plant (CGP3). Albemarle now expects full-year sales volume of 225–235kt LCE, flat to down 4%, with better Wodgina output making up part of the gap. Remediation and ramp-up are due to finish in H2 2026 [8-K Q2; 10-Q Q2].
4. New CEO from BHP. On 2026-09-02 the board named Ragnar Udd, BHP's Chief Commercial Officer and its former President Americas (copper and potash), as President and CEO. He starts 2027-02-01 or earlier by agreement. Kent Masters moves to Executive Chair through the 2027 annual meeting [8-K 2026-09-03]. Two other appointments point the same way. Eduardo Bartolomeo, CEO of Vale from 2019 to 2024, joined the board on 2026-07-21 [8-K 2026-07-23]. Max Hood, previously co-CFO of ODP, became Chief Accounting Officer on 2026-08-24, after the previous CAO retired on 2026-06-01 [8-K 2026-06-05; 8-K 2026-08-21].
5. Balance sheet: net debt nearly gone. Cash was $1.63B against total debt of $1.88B at 2026-06-30. Liquidity was ~$3.2B, and credit-agreement leverage was 0.5×, down from ~1.0× in Q1 [8-K Q2].
6. Share price: ~$170 (2026-06-12) → $116.80 (2026-09-22), −31%. The stock rallied after the Q2 beat, then fell about 25% in a month. Sources give these steps: −5.9% to $133.18 on 2026-08-25 after JPMorgan cut its target to $140; −4.1% to $132.16 on 2026-09-03, the day the CEO change was announced; and $110.91 on 2026-09-18. The lithium price did not follow. Battery-grade carbonate was ~$19.8/kg on 2026-09-24, about 12% below its May peak and still more than double its level a year earlier. The stock fell on fears about where lithium is going (China EVEnterprise valueWhat it would cost to buy the whole company: its market value plus its debt, minus the cash you would get with it. Often very different from market cap. orders, restarts, inventory), not on where it is.
7. Chile labour dispute: resolved at the last minute. On 2026-08-31, 466 of 478 eligible workers voted to strike from 2026-09-02. A preliminary wage deal meeting the union's demands was reached on the last day of mediation and put to a ratification vote. The ratification result was not confirmed in sources this session. It is carried as an open item.
8. June dossier errors, corrected here.

• (a) The conversion dilution was understated. June said "~12M+ shares (≈10%)". The 2.2998M mandatory convertible preferred shares convert on 2027-03-01 at 7.618–9.140 common shares each, or 17.5M–21.0M new shares (15–18% of the 118M basic count) [10-Q Q2, Note on equity]. The rate is set by the 20-day VWAP before conversion. Above ~$131 the minimum 17.5M applies. At ~$117 it is ~19.7M, and below ~$109 it is the maximum 21.0M. The lower the stock, the bigger the dilution. Albemarle's own FY26 guide uses ~136M diluted shares [8-K Q2]. The offset is that ~$167M a year of preferred dividends stops at conversion.
• (b) The FCPA dates. The settlement was finalised in September 2023 (not November) for $218.5M. The DOJ non-prosecution agreement was concluded early, in April 2025, because its terms had been met [10-K FY25, Risk Factors].
• (c) Customers. No customer was more than 10% of sales in FY2025 or FY2024. One Energy Storage customer was ~12% in FY2023 [10-K FY25, Note 25].
• (d) Kemerton Train 1 was put into care and maintenance in February 2026, not 2025. A further $70–90M of decommissioning charges is expected through 2027 [10-Q Q2, Note 10].

What held. The asset thesis is unchanged. Albemarle still owns the lowest-cost Western lithium: Atacama brine plus 49% of Greenbushes. Its own sensitivity table still shows EBITDA moving ~4.5× between $10/kg and $30/kg [8-K Q2]. Specialties again cushioned the business, growing even as the Middle East conflict pushed up bromine costs [8-K Q2].

Phase A — Understand the business

Lens 1 · Overview: carried, see the previous dossier. Updated facts from the filings:

• Two reportable segments since Q1 2026, Energy Storage and Specialties. The retained Performance Catalyst Solutions business and Albemarle's 49% of Refining Solutions (now equity income) sit in "Corporate and all other" [10-Q Q2].
• ~7,800 employees at 2025 year-end, including consolidated JVs: 40% in the Americas, 33% Asia-Pacific, 19% Europe, 8% Middle East and other. About 26% are unionised [10-K FY25].
• The 10-K lists Tesla among competitors in lithium compounds, next to SQM, Tianqi, Ganfeng, Rio Tinto and Pilbara [10-K FY25, Item 1].

Lens 2 · Supply chain: carried. One addition: the Greenbushes CGP3 fire (2026-06-09) is the first operational hit to the crown-jewel asset in the dossier's history. The cost was limited because Wodgina, a 50/50 JV with Mineral Resources, ran better than planned [8-K Q2]. Purchases from related unconsolidated affiliates, mainly Windfield/Talison, were $322.5M in Q2 against $158.6M a year earlier [10-Q Q2]. More of Albemarle's feedstock is now bought from its own JV at higher prices.

Lens 3 · Moats: carried, unchanged. It is a cost moat, not a pricing moat. Realized price ran ~15% below market in Q2 because of spodumene sales and the three-month lag in long-term contracts. That lag cushions the downside as well as delaying the upside.

Segments (re-run: the mix moved)

Q2 2026 vs Q2 2025 [8-K Q2]

Energy Storage adjusted EBITDA includes Albemarle's share of Windfield (Talison) earnings on a pre-tax basis. The scenario table puts that equity income at $0.6–0.7B for FY26 at $20/kg, which is roughly a third of the segment's EBITDA [8-K Q2]. Consolidated sales understate how much of the lithium economics Albemarle owns.

Geography, FY2025 (by shipment destination) [10-K FY25, Note 25]

Korea and Japan fell from $4.52B combined in FY2023 to $1.15B in FY2025, while China held. Albemarle's lithium now goes to Chinese converters and cell makers far more than to the Korean and Japanese cell makers that drove the 2022 peak. The US tariff exposure is small for the same reason: most China production is sold inside China or Asia [10-Q Q2].

Phase B — Measure performance

Earnings result (latest print: Q2 2026, quarter ended 2026-06-30, reported 2026-08-05)

[8-K Q2; 10-Q Q2]

• Net sales $1,743.3M, +31.1%. Energy Storage pricing +73% while Specialties had price +11% and volume +8%. The comparison still includes Ketjen in Q2-25.
• Gross margin 33.9% (Q2-25: 14.8%). Operating income $452.9M (Q2-25: $47.5M).
• GAAP net income $480.0M; $438.3M to common after $41.7M of preferred dividends. Diluted EPS $3.52; adjusted EPS $3.75 on 136.2M diluted shares, assuming the preferred converts.
• Equity income from JVs $151.6M (Q2-25: $78.3M). This is below the operating line and is mostly Talison.
• Tax rate 21.3% GAAP, 19.1% adjusted. The US valuation allowance from Q4-25 is still in place, so most adjustments carry no tax effect.
• OCF $710M, capex $72M, FCF $638M. OCF conversion was 83%. Management attributes part of it to the timing of a larger Talison dividend and to working-capital benefits it calls non-recurring (see Lens 10).
• Year to date: sales $3,172M, adjusted EBITDA $1,522M, OCF $1,056M, capex $170M, and $1,314M of debt repaid, funded by ~$525M of divestiture proceeds plus cash flow.

Scored against the June dossier's base case ("adjusted EPS to common ~$5–7 for FY26"): H1 alone is $6.70. Lithium held near the $20 case rather than the ~$15 blend June used as its base.

Earnings calls (the tone is still confident, and Q3 is guided down)

No transcript on the shelf. Read from the release and published call highlights (paraphrase-risk):

• Demand is described as strong. Masters said global lithium demand was up ~45% YoY through May, supply additions were lagging, and inventories were tight. Albemarle raised its 2026 stationary-storage battery forecast to 900–1,100 GWh and lifted the low end of its 2030 range to 1,500–2,000 GWh.
• Near-term caution. The CFO guided Q3 sales and EBITDA down sequentially.
• The language has moved on from survival. Q2 used "highest value organic growth opportunities" and "disciplined approach to capital allocation" [8-K Q2]. In 2024 the words were "preserve, reduce, optimize". Growth language is back, but capex is still being cut.
• Cost programme: $100M of run-rate improvements delivered year to date, tracking to the high end of the $100–150M 2026 target [8-K Q2].

Comps: carried, with an updated ALB line

The June peer table stands, and its n/a gaps are still open. ALB's own line, updated:

• Market cap ~$13.8B basic (117.96M × $116.80) or ~$16.1B diluted (137.6M shares, with ~19.7M from conversion at the ~$117 rate).
• EV ~$16.6B = diluted market cap + $1.88B debt − $1.63B cash + $0.26B non-controlling interests.
• EV / TTM adjusted EBITDA ($2.02B, twelve months to 2026-06-30 [8-K Q2]) ≈ 8.2×. On the $20-case FY26 high end ($2.6B) it is ~6.4×; on the $10 case ($0.95B), ~17×.
• A forward P/E of ~10.1 is quoted against a diversified-chemicals industry average of ~16.7.

In June ALB screened at ~18–20× trough EBITDA. It now screens at single-digit trailing EBITDA. The asset is the same, but the earnings have caught up and the price has not.

Stock-price catalysts (moves of 5% or more since June)

throughout; daily closes are aggregator-sourced.

• Early August, Q2 beat: the stock rallied after the 2026-08-05 print and the $3.75 adjusted EPS. The size of the move was not confirmed.
• Late August, lithium pullback plus target cuts: −5.9% to $133.18 on 2026-08-25, after JPMorgan cut estimates and its target to $140. Mizuho also cut its target in September.
• 2026-09-03, CEO succession: −4.1% to $132.16 on a day the S&P rose ~0.5%. The drop is under 5% but notable. Handovers in the middle of a cycle tend to be priced as uncertainty.
• Mid-September, lithium demand worries: China EV-order softness, inventory and restarts. The stock was $110.91 on 2026-09-18 and $116.80 on 2026-09-22. The 52-week range is quoted as $79.30–$221.00.

The pattern changed. From 2021 to June 2026, ALB tracked the lithium spot print. From August to September 2026 it fell ~30% while spot fell ~12% from its May peak and stayed near $20/kg. The market has started pricing the expected lithium deck (restarts and an inventory glut) ahead of the spot price. That is new, and it is the reason for this refresh.

Phase C — Judge people & books

Management (re-run: succession)

• CEO-elect: Ragnar Udd, 54, BHP's Chief Commercial Officer since March 2024 (global sales, marketing, procurement and commodities strategy). He was previously President Americas at BHP, running copper and potash from November 2020 to March 2024 [8-K 2026-09-03].
• Pay [8-K 2026-09-03]: $1.3M base salary; 135% target bonus; $1.4M sign-on cash; $11.0M of make-whole RSUs and PSUs to replace forfeited BHP awards; and $7.5M of 2027 long-term incentive awards. The severance multiple is 2.0× outside a change of control.
• Masters becomes Executive Chair when Udd starts, through the 2027 annual meeting. After that his role goes through the normal nomination review [8-K 2026-09-03].
• Board: Eduardo Bartolomeo (Vale CEO 2019–24) joined on 2026-07-21 and sits on the Audit & Finance and the Safety, Sustainability, Operations & Capital committees [8-K 2026-07-23].
• Finance bench: the CAO retired on 2026-06-01. The CFO, Neal Sheorey, was interim principal accounting officer until Max Hood started as CAO on 2026-08-24 [8-K 2026-06-05; 8-K 2026-08-21].

Read [inference]: a BHP commercial chief as CEO and a former Vale CEO on the board mark a move from a specialty-chemicals culture toward mining-major operating and commercial discipline. Udd's BHP role was selling commodities into China and setting commodities market strategy, which is Albemarle's main open problem: how much of the book to sell on index versus contract. A mining-major team can also be read as positioning for consolidation. Nothing filed supports that, and it is flagged as speculation.

Capital allocation: the dividend has been held at $0.405 a quarter, the $1.3B debt tender is done, and capex has been cut three times. No share buybacks in H1 (none in the cash-flow statement) [8-K Q2].

Forensic red flags (re-run: filings now on the shelf)

[10-Q Q2; 8-K Q2] unless noted.

1. Payables owed to related parties tripled. They rose from $134.4M at 2025 year-end to $445.4M at 2026-06-30, mostly money owed to Windfield for spodumene. That +$311M is a working-capital source that flatters H1 operating cash flow, and it reverses when the payable is settled or prices fall. Management itself calls part of Q2's cash conversion "non-recurring working capital benefits" [8-K Q2]. Flag: judge H2 cash flow net of this unwind.
2. Deferred intra-entity profit in inventory quadrupled. It went from $37.2M to $143.0M [10-Q Q2, Inventories]. Albemarle defers its share of Windfield's profit on spodumene still in its own inventory. That keeps reported equity income conservative now and releases ~$106M into later quarters as the inventory is sold. This is not manipulation, but it does move equity income between quarters, and H2 equity income will get a lift from it even if prices soften.
3. The valuation allowance is still in place. Because of the full US allowance booked in Q4-25, most non-recurring items carry no tax effect. The H1 adjusted tax rate was 13.3%, and the FY guide is (50)% to 30% [8-K Q2]. If profits hold, reversing the allowance would be a large one-off GAAP gain. Any EPS model should use a normalised rate, not the reported one.
4. Restructuring is still running. Kemerton Train 1 charges were $33.2M in H1, with $70–90M more to come through 2027 [10-Q Q2, Note 10]. Consulting fees for the cost programme ($19.0M in Q2) were excluded from adjusted results [8-K Q2]. Excluded charges that recur every quarter are worth watching.
5. Turnover in accounting leadership: the CAO retired, the CFO covered the role for 12 weeks, and an external hire replaced him. This is routine, but internal controls should be watched through the FY26 audit. The Q2 10-Q reports no change in internal control over financial reporting [10-Q Q2, Item 4].
6. Divestiture loss: a $52.7M net loss on selling Refining Solutions and Eurecat (Q1). Proceeds were below carrying value [8-K Q2].

Regulatory findings: no SEC Litigation Releases and no AAERs naming Albemarle, 2021-09-25 to 2026-09-25. The June run's AAER HTTP 500 caveat is cleared. Non-SEC matters: the FCPA resolution (September 2023, $218.5M, DOJ NPA concluded April 2025) is corrected above. The Chile labour dispute is operational, not legal. The 10-Q legal-proceedings item points to Note 8 and names no new material case.

Net forensic read: the accounting is still conservative (allowance kept, JV profit deferred). The quality risk is in cash flow, where the payables build and JV dividend timing make H1 look better than it is. Earnings quality is not the concern.

Phase D — Project & stress-test

Forward projection

The anchor is unchanged: Albemarle's own scenario table [8-K Q2], with FY26 volume of 225–235kt and ~40% of salts volume on long-term agreements.

FY2026 adjusted EPS: H1 is $6.70 actual. For H2, take EBITDA of ~$1.08B (the high end of the $20 case less H1) and subtract ~$0.36B of D&A (guide $0.67B less H1 $0.31B), ~$0.07B of interest and ~$0.08B of Windfield tax add-back. That leaves ~$0.57B pre-tax. At ~20% tax and ~$0.03B of non-controlling interests, adjusted net income is ~$0.43B, or ~$3.1–3.3 per share on 136M diluted shares. FY26 adjusted EPS is therefore ~$9.8–10.0 at a flat $20/kg. If H2 averages ~$17/kg it falls to ~$8.5–9. This compares with June's base of ~$5–7.

FY2027: diluted shares move to ~135.5–139M at conversion (Lens 12), and the preferred dividend stops. If lithium averages ~$18–20/kg with modest volume growth once CGP3 ramps, adjusted EBITDA would be ~$2.2–2.6B and adjusted EPS ~$8–10. No consensus line is sourced; the Zacks forward P/E implies roughly $11–12 of forward EPS, but the basis is unclear, so it is not adopted.

**Valuation scenarios **, using equity = EBITDA × multiple − net debt − non-controlling interests, divided by diluted shares:

What today's price implies: at ~8.2× trailing EBITDA the market is capitalising roughly $1.8–2.1B of sustainable EBITDA, which on Albemarle's own table corresponds to ~$15–17/kg. Spot is ~$19.8. The market is pricing a ~15–25% fall in lithium that has not happened yet.

The June 3-year framing (a 1-year cyclical to 3-year structural timeframe) is kept. The June "price target" used consensus (~$213), not the dossier's own arithmetic. This refresh replaces it with the scenario table above. our model create was not run (unattended).

Bull vs bear

Bull. (1) The operating leverage showed up on schedule. Q2 made a 49% EBITDA margin at $19.5/kg, with net debt close to zero. (2) The stock has fallen ahead of the lithium price. The market is pricing ~$15–17/kg with spot at ~$20, while demand grew ~45% YoY through May and Albemarle raised its storage-battery forecast. (3) Things Albemarle controls keep improving: Specialties guidance up, capex down again, the cost programme at the top of its target, and JV profit deferred into later quarters. (4) New management with commodity-trading experience could get more for the resource base. Albemarle's realized price runs ~15% below market, and closing part of that gap would add EBITDA at any lithium price [inference].

Bear. (1) The market may be right about the direction of lithium. China EV orders are slowing, inventories are building, and Australian restarts are ramping. Q3 is already guided down. (2) Cash flow will look worse in H2 as the $311M related-party payable unwinds and dividend timing normalises. (3) Dilution grows as the stock falls: below ~$109 the conversion issues the maximum 21.0M shares. (4) Chile keeps taking a larger share. Q2 EBITDA was "partially offset by increased commission expenses in Chile resulting from the higher pricing" [10-Q Q2, MD&A]. The sliding CORFO royalty takes more of each extra dollar as prices rise. The union also won a better wage deal this round. (5) A CEO change in the middle of a cycle. Udd starts in February 2027. New CEOs often reset guidance or take charges in their first year (a "kitchen-sink" year) [inference].

Pre-mortem (18 months out, the thesis failed): lithium fell to $12–14/kg in 2027 as restarts and inventory met a Chinese EV slowdown. FY27 EBITDA landed around $1.3B. The conversion issued the maximum 21M shares at ~$100. The new CEO took Kemerton and Specialties charges in his first year. The stock sat at $80–95.

Devil's advocate (short)

• "The stock fell while lithium held" can also mean the stock is ahead of the price. Equity markets often move before commodity turns show up in spot prices [inference; not tested here]. If the market is right, today's ~$20 spot is the top, and the Q3 guide-down is the first sign.
• H1 was flattered twice: by working capital (payables) and by pricing at the top of the range. Neither repeats.
• The size of the dilution depends on where the stock trades in February 2027, the month Udd starts. The conversion VWAP window opens about 21 trading days before 2027-03-01. Any drop into that window raises the share count permanently.
• What would prove the bull wrong: a Q3 print (~early November) with realized price below ~$17/kg together with an H2 cash conversion below 60%.

Management questions (updated; the 15 June questions still stand)

New, ordered by how much the answer would tell us:

1. How much of H1 operating cash flow came from the related-party payable build ($134M → $445M), and when does it unwind?
2. When does the $143M of deferred intra-entity profit release, and how much reaches Q3 and Q4 equity income?
3. Where did the CGP3 fire leave Talison's output, and what is the insurance and business-interruption position?
4. What wage and incentive terms were agreed in Chile, and what do they add to Atacama unit cost per tonne?
5. What does the Q3 sequential guide assume for realized price in $/kg, and how much of the drop is volume versus price?
6. Will Ragnar Udd review the ~40% long-term-contract share, or the index-versus-contract mix?
7. Is anything planned before conversion to manage the 17.5–21.0M share issuance, such as a buyback or an early settlement?