Samsung SDI

Phase A — Understand the business

Company Overview

Samsung SDI is South Korea's smallest of the three major battery makers and an affiliate of the Samsung group. The business is two unequal halves:

• Energy Solutions (~93% of revenue) — 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. batteries (prismatic NCA, plus a new 46-series cylindrical line and an emerging LFP line), ESS (grid-scale energy storage), and small-format Li-ion (power tools, e-bikes, IT). The company does not break out EV vs ESS vs small-cell inside this segment — a real disclosure handicap for anyone trying to model the ESS pivot.
• Electronic Materials (~7% of revenue) — semiconductor and display materials (e.g. OLED, semiconductor process materials). Small but the only consistently profitable unit: ~KRW 883B revenue / KRW 130B operating profit in FY2025.

FY2025 group revenue was KRW 13.27 trillion (~$9.6B ``) with a group operating loss of KRW 1.72 trillion.

Customers / contract structure. Battery supply is OEM-qualified, multi-year, take-or-pay-ish volume commitments — but several anchor contracts run through captive JVs rather than arm's-length supply:

• BMW — 46-series cylindrical cells from the Hungary (Göd) plant.
• Mercedes-Benz, Hyundai — prismatic. Hyundai signed a 2026–2032 seven-year prismatic (P6) supply deal for Europe-bound EVs.
• Stellantis — via the StarPlus Energy JV (two US plants, Indiana/Kokomo); a 2nd 34 GWh facility targeted for 2027.
• General Motors — a $3.5B JV (New Carlisle, Indiana), 27 GWh prismatic NCA, production pushed to 2027.
• US ESS customers — multiple multi-trillion-won utility-ESS awards (a >$1B cell deal and a separate >KRW 2T deal disclosed); a Tesla ESS supply talk was reported.

Plain-terms model: a high-fixed-cost cell manufacturer selling into two demand pools — EV (cyclically depressed, structurally Chinese-pressured) and ESS (structurally accelerating, US-policy-favored) — currently pivoting capacity and capital from the former to the latter while sitting on a deep operating loss.

Supply Chain

Upstream → Samsung SDI → end customer, named:

Upstream (inputs):

• Cathode active material (CAM): NCA/NCM from group affiliate-adjacent and Korean suppliers; the rights-issue earmarked ~KRW 500B to build precursor (pCAM) lines in Korea to reduce dependence. For LFP/ESS it signed an LFP cathode supply deal to de-risk North American ESS — a chokepoint, because LFP cathode supply is overwhelmingly Chinese.
• Anode: graphite (China-dominated globally — a structural single-geography chokepoint for all non-Chinese cell makers), plus its "anode-less" solid-state design which removes the graphite anode entirely (a strategic chokepoint-bypass if it scales).
• Separator / electrolyte: Korean and Japanese suppliers; solid-state shifts this to proprietary solid electrolyte (sulfide pathway).
• Equipment: Korean cell-line equipment makers.

The company (cells/packs): plants in Korea (Cheonan, Ulsan, Suwon S-Line pilot), Hungary (Göd), Malaysia (cylindrical), and the US (StarPlus/Indiana, GM/New Carlisle).

Downstream (buyers): BMW, Mercedes, Hyundai, Stellantis, GM (EV); US utilities + data-center / grid integrators + reportedly Tesla (ESS); IT/power-tool OEMs (small cell).

Chokepoints / single-source dependencies:

1. Graphite + LFP cathode = China. Even as Samsung SDI localizes US cell assembly, the materials upstream remain China-centric — the AAM (active anode material) import question even reached the USITC.
2. Captive-JV concentration. GM and Stellantis aren't just customers — they're JV partners; a partner's EV-plan cut hits both the offtake and the capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. commitment.
3. Solid-state electrolyte is proprietary/pilot — a single-source internal dependency with no proven mass-scale yield yet.

(Names provided; this lens does not stay generic.)

Competitive Advantages (moats)

Where the moat is real:

• Premium prismatic/cylindrical quality + safety reputation. Samsung SDI has historically positioned at the high-nickel, high-energy-density premium end — its battery won an InterBattery award for highest energy density. Premium OEMs (BMW, Mercedes) pay for safety/quality track record; that's a genuine switching-cost/qualification moat (12–24 month requalification to switch a cell vendor).
• Solid-state lead. The S-Line pilot (Suwon) delivered first ASB cells to OEMs for 6-month validation; 900 Wh/L / 500 Wh/kg, anode-less, 2027 mass-production target — timeline-matched to Toyota and ahead of most Western peers [KB entity page; web: electrive/Nikkei, 2024]. If solid-state commercializes on schedule, it's a durable IP/process moat at the premium end.
• AMPC subsidy scale in the US — local production unlocks Advanced Manufacturing Production Credits (KRW 275B in FY2025, 3× FY2024). A policy moat against Chinese imports, though policy-dependent.

Where the moat is weak / eroding:

• No cost moat. This is the core problem: Samsung SDI is a price-taker against Chinese LFP. CATL/BYD run near-full utilization, spreading fixed cost and crushing per-unit price. Samsung SDI's premium NCA chemistry is more expensive exactly as the market pivots to cheap LFP.
• Bargaining power is thin. As the #9 EV-cell maker at 2.4% share, it needs its OEM partners more than they need it — the opposite of CATL. ESS gives slightly better pricing power right now only because demand is outrunning non-Chinese supply.

Net: the moat is premium-quality + solid-state optionality + US-policy shelter — not cost, not scale. That's a narrow, conditional moat.

Segments

our figures is empty — **no segment data exists.** All figures.

Quarterly trajectory (the actual story):

• Q4 2025: Battery revenue KRW 3.62T (+28.4% QoQ, +1.6% YoY); battery op loss −KRW 338.5B; ESS at record quarterly revenue; AMPC + EV-volume compensation narrowed the loss.
• Q1 2026: Group revenue KRW 3.58T (+12.6% YoY); group op loss −KRW 155.6B (narrowed 64.2% YoY); net profit +KRW 56.1B (back in the black at the bottom line). Battery: KRW 3.35T rev / −KRW 176.6B op loss. Electronic Materials: KRW 222B / +KRW 21B.

Geography: not cleanly disclosed by segment; the strategic vector is US-ward (StarPlus Indiana NCA-ESS now, LFP-ESS mass production Q4 2026, ~30 GWh US ESS capacity by end-2026) and Europe (Hungary 46-series + prismatic).

The segment read is unambiguous: EV is the anchor dragging the P&L; ESS is the engine the bull case rides; Electronic Materials is a quiet profitable ballast nobody pays for.

---

Phase B — Measure performance

Earnings Result (latest print — Q1 2026, reported 28 Apr 2026)

• Revenue: KRW 3.58T, +12.6% YoY. Annualizes to ~KRW 14.3T `` — above FY2025's KRW 13.27T, i.e. the top line has inflected up.
• Operating loss: −KRW 155.6B, narrowed 64.2% YoY. Still a loss, but the slope is the point.
• Net income: +KRW 56.1B — returned to profitability at the net line. (Net positive while operating negative implies below-the-line help — equity-method income from the Display stake and/or FX; worth flagging as low-quality earnings.)
• Drivers: demand recovery across ESS, UPS, BBU (battery backup for data centers), power tools; "significant project awards for utility ESS from major US customers". This is the AI/data-center power story showing up in the numbers (Korea Herald headline: "Samsung SDI narrows loss on AI-driven ESS demand").
• Margin: still negative at the operating line for batteries (−KRW 176.6B on KRW 3.35T ≈ −5.3% battery operating margin ), offset by Electronic Materials' ~+9.5% margin .
• Guidance/tone: management guided gradual recovery from Q2 2026, ~50% increase in ESS battery sales in 2026, and return to profitability by end-2026.
• Balance-sheet flags: FCFFree cash flowCash left after paying to run and maintain the business. Unlike profit, it is hard to flatter with accounting choices. guided negative through FY25–FY27 on heavy capex; no dividend FY25–27. Year-end 2024 cash ~KRW 2.1T, net debt/EBITDA ~0.4×, but FY2025's KRW 1.72T operating loss + KRW 3.3T capex means leverage has risen materially since — hence the rights issue and Display-stake sale.

Unusual vs. own history: a net profit alongside an operating loss is atypical and equity-method/FX-flattered — do not read the "return to profit" headline as operational.

Earnings Calls (sentiment trend)

No transcripts/ on disk — all ``. Tracking the last ~4 quarters of management messaging:

• Q4 2024 / early 2025: defensive — "EV slowdown," "challenging demand," capex discipline. The rights-offering (Mar 2025) landed into a falling stock and shareholder backlash — tone was apologetic/justifying.
• Q2–Q3 2025: transition language — "ESS momentum," "US localization," "AMPC benefit." Began re-pointing the narrative from EV-defense to ESS-offense.
• Q4 2025: "record ESS quarterly revenue," "narrowing losses," AMPC tripling.
• Q1 2026 call: explicitly "navigating challenges with strategic growth initiatives," ESS/UPS/BBU demand recovery, 50% ESS sales growth target, profitability by year-end.

Shift: the vocabulary has rotated from "EV demand is weak, please bear with the dilution" (2025) → "ESS and data-center storage are pulling us back to profit" (2026). They stopped leading with EV volume and started leading with grid-scale/AI-storage and US policy. That tonal rotation is the single clearest signal in the file — management has repositioned the equity story from cyclical-EV-victim to structural-ESS-winner.

Comps

Read: The entire Korean cohort (Samsung SDI, LGES) is loss-making on EV, so P/E is meaningless and the group trades on P/S + recovery optionality + capacity value, not earnings. Samsung SDI at ~1.7× P/S vs LGES's larger cap reflects its smaller scale and #9 position. CATL is the only profitable scaled peer and the valuation anchor that matters — it is bigger than the entire Korean trio combined and profitable, which is the whole bear case in one row. Precise current EV/EBITDA for the Chinese names was not sourceable in this pass; flagged rather than fabricated.

Stock-Price Catalysts (>5% moves, last ~5y pattern)

Mostly ``:

• Mar 2025 — rights offering announcement: stock dropped on KRW 2T dilution (16.8% share increase); shareholder backlash. Market reacts to dilution.
• 2024–early 2025 — EV-demand downgrades / "EV winter": persistent de-rating; the stock fell toward its 52-week low of ₩165,900. Market reacts to EV cyclicality.
• Late 2025–2026 — ESS / data-center demand + US supply deals: the >$1B and >KRW 2T US ESS awards, AMPC tripling, and Q1'26 net-profit return drove a violent re-rating — +42.6% YTD 2026, to a 52-week high ₩723,000. Market now reacts to ESS/AI-storage order flow.
• Feb 2026 — Samsung Display stake-sale plan (~KRW 10T): read as a balance-sheet fix funding the ESS pivot — supportive.

Pattern: for ~3 years the tape traded Samsung SDI as an EV-cycle + dilution story (down). In 2026 the reaction function flipped to ESS/data-center order momentum + US policy (up, hard). The market has effectively re-cast the name from "EV cell laggard" to "AI-power-infrastructure storage play." That re-rating is now the risk: a lot of the turnaround is priced.

---

Phase C — Judge people & books

Management

• CEO: Leadership transitioned — Yoon-Ho Choi was President/CEO 2022–2024; recent (2026) reporting references Choi Joo-sun as CEO, brought in to strengthen the battery supplier ecosystem. (Exact current-title confirmation is murky in public sources — flag as ``, medium confidence.)
• Track record: Samsung SDI under the prior regime built the premium-OEM franchise (BMW/Mercedes) and the solid-state pilot lead — genuine technical delivery. But it also lost EV share (to #9) and swung from profit to a KRW 1.85T Energy-Solutions loss — a real operational failure on the cost/scale axis.
• Capital allocation — the crux: three big moves in ~15 months:

1. KRW 2T rights offering (Mar 2025) — dilutive (16.8% more shares), backlash; funded Hungary lines, the GM JV, and Korean precursor lines.
2. Samsung Display stake sale (~KRW 10T / 15.22%, Feb 2026) — monetizing a passive holding (~18% of current market capMarket capitalisationThe share price multiplied by the number of shares. What the market says the equity is worth. ``) to fund ESS/GM-JV/solid-state and repair the balance sheet. This is the smart move — selling a non-core financial asset rather than diluting further.
3. Capex discipline — FY2025 capex cut to KRW 3.3T (down YoY), redirected toward ESS/US.

• Skin in the game / insider ownership: Samsung-group affiliate structure; no our figures — n/a.
• Red flags: (1) dilution then asset-sale in quick succession signals the balance sheet was tighter than the headline net-debt/EBITDA implied; (2) net profit while operating-loss (equity-method/FX flattering) risks a "return to profit" narrative that isn't operational; (3) governance — capital decisions inside a chaebol can favor group interests over minority holders (the Display stake is an intra-group asset).
• Archetype: professional managers inside a chaebol, not founders. Implication: disciplined execution and group backing (a real downside-support), but slower, more politically-constrained capital decisions and minority-holder-alignment risk.

Forensic Red Flags

our figures empty — all /, no filing-grounded numbers.

• Earnings quality (highest-priority flag): Q1 2026 net profit (+KRW 56.1B) sits on top of an operating loss (−KRW 155.6B). The gap is bridged below the operating line — plausibly equity-method income (Samsung Display stake) and/or FX. Headline "returned to profit" is therefore low-quality; the operational business still lost money. Watch whether the Display-stake sale removes that equity-method income going forward — selling the stake fixes the balance sheet but removes a non-operating earnings prop.
• AMPC dependence: KRW 275B of FY2025 "other operating income" is US production tax credit, not commercial margin. It flatters Energy Solutions and is policy-contingent — any IRA/AMPC rollback would re-widen the loss. Segment operating income is materially subsidy-supported.
• Capitalized capex vs negative FCF: heavy capex (KRW 3.3T) with guided-negative FCF through FY27 means rising PP&E and the risk of future impairment if US/EV volumes (GM, Stellantis) underdeliver — JV-plant assets are the impairment-exposed line.
• Inventory/receivables: not disclosed in this web pass — n/a. (Korean battery makers carried elevated inventory through the EV-destocking of 2024–25; flag to verify against the audited statements.)
• JV / related-party complexity: StarPlus (Stellantis), GM JV, and the Samsung Display equity stake are all related-party-flavored — earnings and asset values flow through entities not fully consolidated/transparent in summary disclosure.

Regulatory findings (required sub-section).

• SEC (EDGAR LR/AAER): None possible — Samsung SDI has no CIK and is not an SEC filer. Zero SEC findings by construction, not by clean record.
• Non-SEC web search ("Samsung SDI" (FTC/DOJ/FDA/CFPB/consent decree/settlement/fine/penalty) enforcement): no material enforcement action surfaced in this pass. The most relevant regulatory event is the USITC review of Chinese AAM (anode active material) imports — a trade matter that went in favor of US producers, i.e. industry-protective, not an action against Samsung SDI. n/a — no material enforcement found .
• Item 3 Legal Proceedings: no 10-K exists (foreign private issuer, no EDGAR) — Korean equivalent (사업보고서) not ingested. n/a — not in research layer.
• Verdict: No material regulatory or legal findings surfaced — verified via SEC EDGAR EFTS (returned N/A, no CIK) and web search as of 2026-06-18; the company's own Korean disclosure was not machine-read this pass and should be checked before any high-conviction position.

---

Phase D — Project & stress-test

Forward Projection

No consensus EPS was sourceable in this pass (Korean sell-side estimates not retrieved) — so the path below is `` from disclosed actuals + guidance, not a market consensus, and EPS is expressed directionally because the share count just changed (rights issue).

Anchors: FY2025 group revenue KRW 13.27T, operating loss −KRW 1.72T; Q1'26 run-rate ~KRW 14.3T revenue ; management guides **+50% ESS sales** and **profitability by end-2026**; ~**82.2M shares** post-rights-issue .

EPS: n/a as a consensus number; directionally, FY2026 GAAP EPS likely still negative or marginally positive (equity-method-flattered), with the first clean operating-EPS year FY2027 base / FY2026 bull. The honest statement: this is a turnaround being priced on the slope of the ESS ramp and the credibility of "profit by end-2026," not on a defensible forward P/E.

(Per --watchlist rules: no our model create logged in this loop.)

Bull vs Bear

Bull case. Samsung SDI is the cleanest non-Chinese, US-localized ESS supplier at the exact moment AI/data-center power demand turns grid-scale storage into a structural growth market. It has: real US capacity (StarPlus Indiana NCA-ESS live; ~30 GWh US ESS by end-2026; LFP-ESS mass-production Q4'26), AMPC subsidy insulation from Chinese imports, multi-trillion-won US ESS order book (incl. reported Tesla talks), a premium-OEM EV franchise (BMW/Mercedes/Hyundai/GM/Stellantis) that survives the cycle, a 2027 solid-state lead (900 Wh/L, anode-less) that could re-open a high-margin premium niche, and a fortified balance sheet from the ~KRW 10T Display-stake sale that funds all of it without further dilution. The Q1'26 net-profit return + 64% loss-narrowing + 50% ESS-growth guide is the inflection; +42.6% YTD says the market agrees the trough is in.

Bear case (permanent-impairment risks):

1. Chinese LFP deflation is structural, not cyclical. CATL (39%+ share, profitable, >$140B) and BYD run near-full utilization and are now pushing LFP into ESS — Samsung SDI's escape hatch. China warned 16 firms against below-cost price wars because the overcapacity is that severe. If LFP-ESS prices deflate the way LFP-EV did, the ESS margin thesis compresses before it scales.
2. Subsidy + below-operating-line dependence. Strip AMPC (KRW 275B) and equity-method income and the operational business is deeply loss-making. Both props are policy/asset-sale-contingent and shrink once the Display stake is sold.
3. JV/offtake concentration. GM and Stellantis EV plans have already slipped (GM JV pushed to 2027); a Western OEM EV-plan cut hits offtake and JV capex commitments simultaneously.

Pre-mortem (18 months out, thesis broke): It's late 2027. US ESS pricing rolled over as Chinese LFP cells (via third countries / domesticated AAM) undercut on cost; AMPC was trimmed in a budget fight; the GM/Stellantis JVs ramped slow and an impairment hit the Indiana assets; solid-state slipped to 2028; the Display-stake cash got absorbed by the operating loss instead of growth. The +42.6% re-rating round-tripped.

Are multiples too high? On earnings, there are none (loss-making) — so the stock is a recovery option priced at ~1.7× P/S that already moved +42.6% YTD. The re-rating has pulled forward a lot of the turnaround; the asymmetry is no longer cheap.

Contrarian view (what the market is refusing to see): Either direction. The bear-contrarian: the market is treating an AMPC- and equity-method-flattered "return to profit" as operational, and is paying for an ESS-margin durability that Chinese LFP will not allow. The bull-contrarian: the market still anchors on the dead EV-share number (#9, 2.4%) and under-prices the Electronic Materials profit ballast + the 2027 solid-state call option, which are worth more than a 2.4%-share EV laggard implies.

Devil's Advocate (short-seller)

Dismantling the bull case:

• Structural break in how it makes money: it's a high-cost premium-chemistry cell maker in a market deflating toward cheap LFP it doesn't lead. The ESS pivot doesn't escape this — it just moves the cost disadvantage to a new product line where CATL/BYD/EVE are already pushing LFP at scale.
• Revenue concentration: EV revenue funnels through a handful of Western OEM JVs (GM, Stellantis) whose EV ambitions are being cut, not raised; ESS concentrates in a few large US utility/Tesla awards — lose one and the "50% growth" guide breaks.
• Moat weaker than bulls think: premium-quality is a 12–24mo switching cost, not a cost or scale moat — and it's the wrong moat for a price-deflating commodity market. Solid-state is pilot-stage with unproven mass yield; "2027" has been the target since 2023.
• Most dangerous underestimated competitor: not CATL (obvious) — it's the Chinese LFP-ESS cohort (EVE, CATL's ESS arm) routing around US tariffs via domesticated AAM, plus LGES, which is bigger and also chasing US ESS with AMPC.
• Worst capital-allocation / accounting tells: rights issue then a KRW 10T intra-group asset sale in 15 months (balance sheet tighter than disclosed); net profit propped below the operating line by the very stake it's about to sell; AMPC booked as operating income.
• Assumptions that must hold for today's price: ESS +50%/yr at non-deflating prices; AMPC intact; GM/Stellantis ramp on time; solid-state on schedule; Display cash funds growth not losses. That's a stacked conditional.
• −20–30% growth-disappointment scenario: if ESS growth halves and EV stays soft, FY2026 op. loss stays near FY2025's ~−KRW 1.7T, the "profit by year-end" guide fails, AMPC can't cover it, and a +42.6%-YTD stock with no earnings support de-rates hard.
• Single permanent-impairment scenario (and plausibility): Chinese LFP-ESS deflation + a US OEM JV cut + AMPC trim → write-down of US JV assets and a structurally sub-scale #9 cell maker that can't earn its cost of capital. Plausibility: moderate — the China-deflation leg is already underway; the question is whether ESS demand growth outruns it.

Management Questions (ordered by information value)

1. ESS pricing is the whole bull case — what are your contracted ESS cell ASPs and gross margins today vs. 12 months ago, and how much have they already compressed against Chinese LFP?
2. Of the Q1'26 net profit, how much was equity-method income from the Samsung Display stake? Once that stake is sold, what is the pro-forma operating-and-net path without it?
3. What is the break-even utilization and timeline for the GM and StarPlus JVs, and what contractual protection do you have if a partner cuts EV volumes again?
4. Quantify AMPC dependence: at what AMPC level does Energy Solutions reach operating breakeven without the credit?
5. Solid-state has targeted 2027 since 2023 — what is the current yield on the S-Line, and what specifically gates mass production now?
6. The ~KRW 10T Display proceeds — exact split between (a) debt reduction, (b) ESS/US capex, (c) solid-state — and how much is a buffer against further operating losses?
7. With LFP-ESS the growth vector, how do you win on cost against CATL/EVE given your higher-cost manufacturing base and Chinese cathode-supply dependence?
8. What is your path off Chinese graphite and LFP cathode — does anode-less solid-state materially reduce China input dependence, and by when?
9. After the rights issue and the asset sale, what is peak net-debt/EBITDA through the trough, and your covenant headroom?
10. EV share fell to #9 / 2.4% — is EV now a managed-decline business funding the ESS/solid-state pivot, or do you intend to regain share, and how?
11. What specific US data-center / grid customers underpin the +50% ESS guide, and how concentrated is that order book?
12. How exposed is the ESS thesis to AMPC/IRA rollback under shifting US policy, and what is the geographic hedge?
13. Electronic Materials is your only profitable unit — is it core, or a divestiture candidate, and why isn't it more prominent in the equity story?
14. What is your capacity-conversion plan — how much idled/under-utilized EV line can convert to ESS/LFP, at what cost and speed?
15. Define the 2026 "return to profitability" precisely: operating or net, with or without AMPC and equity-method income?

---