LG Electronics

Phase A — Understand the business

Company Overview

The business in plain terms. LG Electronics makes and sells physical goods across four operating "Companies" (its 2025 segment structure), and is grafting recurring/services revenue onto that hardware base. It is a subsidiary of the LG chaebol: LG Corp holds ~33.7–35.1% of common shares, and the Koo family controls LG Corp (Chairman Koo Kwang-mo the largest individual holder). It is not the same entity as LG Energy Solution (batteries, separately listed) or LG Display (panels) or LG Chem — though it owns equity stakes in the affiliates (see Lens 4/10).

The four segments (renamed in 2025 — old H&A/HE/VS/BS → new HS/MS/VS/ES):

• HS — Home Appliance Solution (fridges, washers, cooking, air care): FY2025 revenue KRW 26.13T, OP KRW 1.28T (~4.9% margin) — record revenue. The profit engine.
• MS — Media Entertainment Solution (TVs, monitors, webOS platform): FY2025 revenue KRW 19.43T, operating LOSS KRW 750.9B. The problem child — Chinese TV competition (Lens 3/8).
• VS — Vehicle Solution (in-vehicle infotainment, telematics, 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. powertrain components, ZKW lighting): FY2025 revenue KRW 11.14T, OP KRW 559B — record highs on both. Order-backlog-driven; a decade-long turnaround management says is "on the highway."
• ES — Eco Solution (HVAC / chillers / air solutions, incl. the data-center-cooling push): FY2025 revenue KRW 9.32T, OP KRW 647.3B (~6.9% margin). The designated B2B growth driver.

Consolidated FY2025: revenue KRW 89.2T (record, 2nd year running), operating profit KRW 2.48T — OP down YoY (from KRW 3.42T in FY2024) on TV weakness + heavier marketing. Blended OP margin ~2.8% consolidated, but that masks a healthy ~7% appliance business dragged by the TV loss and corporate/eliminations.

Contract structure / payment terms. Overwhelmingly transactional hardware sale (retail + channel) — not take-or-pay, not contracted recurring. The strategic thrust is precisely to change this: (1) subscription/rental (hardware + care bundled into monthly plans) reached ~KRW 2T in 2024, +75% YoY, +29% in FY2025 to ~KRW 2.5T; (2) webOS ad/content (LG Channels, 3,800+ FAST channels, 29 countries) passed KRW 1T in 2024, ~4× vs 2021; (3) VS backlog gives multi-year revenue visibility on auto components. Customer concentration is low on the consumer side (millions of end buyers via retail); higher on VS (a handful of global automakers) and emerging on ES (hyperscaler/data-center buyers).

Supply Chain

Named stakeholders along the chain (names or it didn't happen):

Upstream inputs → LGE:

• Displays/panels: LG Display (LGE owns 36.7%, equity-method ) for OLED TV panels; also buys from Samsung Display (QD-OLED) and Chinese makers (BOE, CSOT/TCL) for LCD.
• Semiconductors/components: LG Innotek (LGE owns ~30.08% ) for camera modules, substrates, sensors; plus merchant chips from the broad supply base (MCUs, power, memory).
• Compressors/motors: largely in-house (LGE's core appliance IP — inverter linear compressors, motors), a genuine vertical-integration moat in appliances and now leveraged into HVAC chillers.
• Batteries/power: LG Energy Solution (affiliate) for EV-related and, per the 2026 Nvidia MOU, 800V DC data-center power development.
• Metals/plastics/logistics: commodity steel, resin, rare-earth magnets; global freight — the input that tariff policy is now re-routing (Lens 8).

LGE (manufacturing footprint): Korea (Changwon appliance, Pyeongtaek chiller plant + the "AI Data Center HVAC Solution Lab" testbed ), plus plants in the US (Clarksville, Tennessee — washers/dryers, ~20% US share; refrigerator/TV lines under consideration), Mexico (refrigerators, TVs — the tariff-exposed node), India (via the newly-listed LG Electronics India), China, Vietnam, Poland (ZKW auto lighting), Brazil.

Downstream → end customer:

• Consumer: big-box + e-commerce (Best Buy, Amazon, Costco, regional retail) → households.
• VS: direct to global automakers (OEM component supply; GM, Hyundai/Kia, VW-group and others via ZKW).
• ES / data-center cooling: the new channel — direct/EPC to hyperscalers and colo operators; the Nvidia MOU aligns LGE's CDUs/cold-plates to Nvidia's DSX reference design, positioning LGE as a thermal-management vendor inside AI-factory builds.

Chokepoints / single-source dependencies: (1) Advanced OLED panel supply concentrated in LG Display + Samsung Display — a strategic dependency that also props up a struggling affiliate LGE part-owns. (2) Mexico manufacturing is a tariff chokepoint for the US market (Lens 8). (3) Rare-earth magnets / China inputs for motors and compressors. (4) In data-center cooling, LGE is the supplier to a chokepoint (Nvidia's ecosystem), which is upside, not dependency — but it makes the ES growth story a derivative bet on the AI-capex cycle holding.

Competitive Advantages (moats)

Where the moat is real:

• Appliance manufacturing + component vertical integration. In-house inverter compressors/motors, decades of process know-how, and a globally recognized premium brand give HS a durable ~7–8% operating margin in a "mature" category — genuinely hard to replicate. This is the crown jewel.
• HVAC/chiller engineering leveraged from appliances. The same compressor/thermal IP underwrites the data-center-cooling entry; LGE isn't a startup in cooling, it's an incumbent air-conditioning maker moving up-stack (chillers → CDUs → direct-to-chip/immersion). The DTC lineup at 1.4 MW capacity and in-house compressors are a real, differentiated asset.
• VS order backlog + qualified-supplier status. Automotive is a slow-moving, design-win, multi-year-qualified business — the decade it took to reach profitability is itself a barrier to the next entrant.
• webOS installed base. Hundreds of millions of LG smart TVs create a two-sided ad platform (LG Channels) with genuine switching friction for content/ad partners — a nascent but real software moat.

Where the moat is weak / eroding:

• TVs (MS). This is a moat loss in progress. LG's global TV unit share fell to ~9% (Nov 2025) while TCL (~12%) and Hisense (~11%) each passed it; TCL+Hisense+Xiaomi (~31.8% in Q3'25) now exceed Samsung+LG (~28.5%) combined. LG still leads premium OLED (~51% share, down from 54% ) but Samsung's QD-OLED is closing (24%→31%). Brand + panel edge is being out-priced.
• Robotics. LGE has no proven moat here yet — Bear Robotics gives a service-robot software stack + fleet-management, and LG CLOi gives hardware, but this competes with a crowded field (Bear vs. peers in hospitality/logistics; humanoids are a different, unproven league). The moat is aspirational.

Bargaining power. Strong over most component suppliers (scale buyer, plus captive affiliates). Weak-to-moderate over consumers in TVs (commoditizing, price-led). Improving in VS (qualified supplier, but automakers are tough). Genuinely rising in data-center cooling if the Nvidia alignment sticks — being spec'd into a reference design is leverage.

Segments

FY2025 by operating Company (all; USD ):

Geography. LGE does not cleanly break OP by geography in the English releases sourced here, so a geographic OP split is n/a. Directionally: the Americas (esp. US, ~20% appliance share) and Korea are the profit core; Mexico manufacturing serves the US; India was material enough to IPO separately (LG Electronics India listed Oct 2025 at up to $8.7B valuation, ~50% pop on debut, the strongest Indian IPO demand since 2008 ) — India is now a partially-listed value stub inside the consolidated entity.

The segment story in one line: appliances pay the bills (HS), TVs bleed (MS), autos and HVAC are the sourced growth (VS+ES combined passed KRW 1T OP for the first time in 2025 ), and platform/subscription/India/robotics are the optionality the stock got excited about.

---

Phase B — Measure performance

Earnings Result (latest print — Q1 2026)

Q1 2026 (reported ~Apr 2026), all:

• Revenue KRW 23.73T — record Q1, +4.3% YoY.
• Operating profit KRW 1.67T — +32.9% YoY, 3rd-highest Q1 ever; OP margin ~7.1%, best in five quarters. LGE beat consensus on margin.
• Segments: HS revenue KRW 6.94T (all-time quarterly high), OP KRW 570B, 8.2% margin despite tariffs/input costs — the appliance engine is intact. MS revenue KRW 5.17T, OP KRW 372B (TV back in the black seasonally). VS revenue KRW 3.1T, OP KRW 212B — record highs. ES revenue KRW 2.82T, OP KRW 249B — down YoY (weak local sentiment / Middle-East drag / higher personnel cost).
• What it says: the operating business is fine-to-good — the margin beat and appliance resilience under tariffs are the real signal. Note the contrast with Q4 2025, when MS/TV drove a net loss for the quarter. LGE's earnings are seasonally and mix-volatile; a single strong Q1 does not underwrite the parabolic YTD stock move.
• Balance-sheet flags (from valuation aggregators): current ratio ~3.30, D/E ~0.49 — comfortable liquidity and moderate leverage. Full FY2025 net income attributable to owners: n/a cleanly (H1 2025 net income to controlling interest was KRW 1.40T, +36% YoY; a clean audited full-year owners' figure wasn't isolable in the sourced material).

Earnings Calls (sentiment trend)

No transcripts on the shelf (transcripts=0); this is web-derived and labeled.

Across the FY2024 → Q1 2026 arc, management (CEO Cho Joo-wan / "William Cho" through most of it) ran a remarkably consistent script:

• What they keep saying: "qualitative growth," "structural competitiveness," B2B (now ~35% of sales), platform/webOS, subscription/D2C, and the "Triple 7" target (7% revenue growth, 7% OP margin, 7× EV/EBITDA) — pulled forward two years to 2028. Tone: confident, transformation-narrative.
• What got louder in 2026: AI / physical AI / data-center cooling / Nvidia. The 2026 strategy (Nasdaq, Jan 2026) foregrounds "AI and B2B focus"; by June the language is a "comprehensive AI alliance" with Nvidia and a CEO-level Robotics Business Center.
• What faded: smartphone legacy (exited 2021) is gone from the story; TV/MS is discussed defensively ("three-pronged strategy to stabilize market share" ) rather than as growth.
• Sentiment read: management sentiment is bullish and increasingly AI-inflected — which is exactly why a skeptic should separate management's (self-interested, promotional) narrative from the sourced financials. The narrative got ahead of the P&L in 2026 (Lens 8/12/13).

Comps

Peer set: global appliance/consumer-electronics majors + Korean large-caps. Multiples are with source/date, or n/a. No multiple is fabricated.

Read of the comps. The clean-comp problem is real: LGE is a conglomerate, so no single peer maps. Two anchors matter. (1) Forward P/E ~9.9 vs trailing ~19.3 implies the market expects a large 2026 earnings step-up (consistent with the FY2026 street model of ~KRW 3.8T OP, +55% ) — i.e., a lot of good news is already in the forward number. (2) Sum-of-the-parts pressure: LGE owns 36.7% of LG Display + ~30% of LG Innotek + a controlling stake in a separately-listed India unit worth ~$8.7B at IPO. That India stake alone is a meaningful fraction of LGE's ~$23B market capMarket capitalisationThe share price multiplied by the number of shares. What the market says the equity is worth. — a classic Korean holdco-discount / SOTP setup where the parent trades below the stubs. That is the sober valuation case, not "robotics multiple."

Stock-Price Catalysts (moves >5%, last ~5 years — 2026 dominates)

The defining event set is 2026, and it is a textbook narrative round-trip:

• Feb–Apr 2026: shares languishing (~KRW 124,200 on Apr 17) — an unloved appliance conglomerate.
• Apr→May 2026 (+~90% in a month): data-center-cooling order momentum + Nvidia "physical AI" collaboration chatter → +9.68% single-day to ~238,000 (mid-May).
• Jun 2, 2026 — the peak: KRW 392,500, up ~329% YTD / "quadrupled" on physical-AI/robotics euphoria and a Koo–Huang meeting.
• Jun 26, 2026: Nvidia partnership formalized/deepened (M.A.P. — Mobility / AI Infra / Physical AI; LG Data Factory; CDUs/cold-plates on Nvidia DSX; LGES 800V DC power).
• Jun–Jul 2026 — the unwind: AI-beneficiary basket corrects hard; LGE and Doosan Robotics halve from peak as pre-reflected expectations get profit-taken. Doosan fell ~50.5% (166,700→82,500); LGE ~185,700 on Jul 6 — roughly −53% from the 392,500 peak yet still up strongly YTD.

What the pattern reveals. For most of its history LGE reacted to fundamentals — quarterly appliance/TV margins, VS order wins, seasonal net swings. In 2026 it became, briefly, a theme stock: it now reacts to (a) Nvidia/AI-capex sentiment, (b) robotics/physical-AI headlines, and (c) data-center-cooling order data-points — far more than to the ~7% appliance margin that actually pays for the company. The stock's beta to the "physical AI" narrative is the single biggest risk and the single biggest source of realized volatility. Anyone taking a position is really taking a view on whether the AI-infra narrative re-inflates or the fundamentals reassert.

---

Phase C — Judge people & books

Management

CEO through the period: Cho Joo-wan ("William Cho"). Career LG since 1987; CEO from Nov 2021, i.e., took the chair right after the smartphone exit. Note a leadership transition: reporting from late 2025/2026 indicates Cho stepped down after ~4 years as part of a generational reshuffle. This is a material governance data-point — the architect of the "qualitative growth"/Triple-7/B2B strategy is handing over as the AI narrative peaks. Confirm the incoming CEO and continuity before sizing any position (open item).

1. Track record. Under Cho: record revenue in 2022/2023/2024/2025; VS turned profitable (a decade in the making) and hit record OP; B2B to ~35% of sales; webOS ads past KRW 1T; subscriptions ~KRW 2–2.5T. Genuine mix-shift execution. The blemish: TVs (MS) lost the volume war and swung to full-year operating loss in 2025 — not solved on his watch.
2. Tenure & skin in the game. Professional manager, not owner. Real skin in the game sits with the Koo family via LG Corp (~33.7% of LGE); the CEO is an agent of that control bloc. Insider-transaction data: n/a (no our figures).
3. Capital allocation. Notable, mostly rational moves: (a) exited money-losing smartphones (2021) — value-accretive discipline; (b) acquired Bear Robotics majority (51%, closed 2025 after a $60M initial stake) — a bet, not yet a return; (c) IPO'd LG Electronics India (Oct 2025) — surfaced value and raised local capital; (d) sharpened shareholder returns 2022–25. ROE/ROIC are low (~4.6%/~4.9% latest ) — the honest read is a well-run operator in structurally low-return end-markets, improving the mix but not yet the through-cycle returns.
4. Red flags (governance). Standard chaebol concerns: cross-shareholdings (LGE holds LG Display + LG Innotek; LG Corp holds LGE) that can subordinate minority-holder interest to group strategy (e.g., propping panel supply). No sourced related-party scandal, but the structure itself is the flag.
5. Founder vs professional. Professional managers running a family-controlled group. Implication: execution-competent, strategically coherent, but ultimate capital-allocation priorities serve the group/family, and the CEO chair changes with the family's reshuffle calendar — as it just did.

Forensic Red Flags

Web-only; no filings on the shelf to tie out. Flags to price:

• Equity-method affiliates flatter/obscure the picture. LGE carries 36.7% of LG Display (book value ~KRW 2.32T ≈ 3.53% of assets at end-2024 ) and ~30% of LG Innotek by the equity method. LG Display has been a chronic loss-maker; equity-method losses can drag net income below operating profit, and the carrying values are exposure to panel-industry cyclicality. Watch for impairment of these stakes.
• Consolidated OP margin (~2.8% FY2025) is depressed by one segment (MS/TV loss). Not an accounting red flag per se, but it means headline consolidated profitability understates the appliance/HVAC/auto core — and equally, any "clean-up" of TV (exit/JV) would step-change reported margins. (Note the recurring, quickly-denied rumor LG might sell/JV the TV business to Hisense — a real strategic question, not just noise.)
• Subscription accounting. Rapidly-growing rental/subscription (KRW ~2.5T) mixes hardware and financing/service — worth confirming revenue-recognition timing and any receivables/financing build as it scales. Not flagged as aggressive by any source; simply the line to watch.
• Preferred shares + India minority. LGE has preferred as well as common shares, and now a listed India minority — consolidated figures include minority interest that dilutes attributable earnings. Use attributable-to-owners not consolidated net when valuing.
• Cash vs earnings / working capital: specific FCFFree cash flowCash left after paying to run and maintain the business. Unlike profit, it is hard to flatter with accounting choices., receivables, and inventory-vs-revenue divergence figures are n/a at line-item granularity here (no research-layer financials). Balance-sheet ratios (current 3.3, D/E 0.49) are reassuring at the aggregate.

Regulatory findings (required sub-section). From regulatory/regulatory-findings.md: total_sec_findings: 0 — LG Electronics has no CIK, so no SEC EDGAR (LR/AAER) search is possible; the file explicitly notes this and defers to web search.

• SEC (EDGAR LR/AAER): none possible / none — not an SEC registrant.
• Non-SEC / web ("LG Electronics" (FTC/DOJ/FDA/CFPB/consent decree/settlement/fine) enforcement): the material recent action is a Texas AG (Ken Paxton) privacy settlement, Dec 2025, over smart-TV ACR viewing-data collection — LG agreed to stop collecting ACR data without informed consent. Otherwise consumer class-actions (LG linear-compressor refrigerators; OLED burn-in; washer drum-bearing) reached settlements in 2024–25. These are product/privacy matters — no sourced antitrust/DOJ/accounting-fraud enforcement against LGE in the window.
• Item 3 (Legal Proceedings) from a 10-K: n/a — LGE files no 10-K (no CIK). The K-IFRS/DART equivalent was not pulled in this web-only pass (open item if a position is taken).
• Bottom line: No material regulatory or accounting-fraud findings surfaced — verified via the regulatory-findings file (SEC N/A, 0 findings) + web search (privacy/product class-actions only) as of 2026-07-06. The genuine "books" risk here is cyclical/affiliate-carrying-value and mix, not fraud.

---

Phase D — Project & stress-test

Forward Projection

Bottom-up from FY2025 actuals + Q1 2026 + street. Output; inputs labeled. No our model create (unattended watchlist rule).

Anchors: FY2025 OP KRW 2.48T on 89.2T revenue. Q1 2026 OP 1.67T, +32.9% YoY, 7.1% margin. Street FY2026: revenue ~KRW 94.33T (+6%), OP ~KRW 3.8T (+55%). Shares ~163M common; TTM P/E ~19.3, fwd ~9.9 — the fwd/trailing gap implies consensus EPS roughly doubling into 2026.

Because a clean audited FY2025 attributable-EPS wasn't sourceable, I project operating profit (better-sourced) and give an EPS band as with explicit assumptions rather than a false-precision number.

EPS band: at ~163M shares and a ~55–65% OP→net-attributable conversion (tax + minorities + equity-method affiliate swing), FY2026 attributable EPS lands very roughly in a KRW 12,000–16,000 range in the base case. Do not treat this as a hard number — the affiliate/minority conversion is the swing factor and is not cleanly sourced. Against ~KRW 185,700, that base EPS implies a forward P/E ~12–15× on operating recovery — reasonable, not cheap, not the "physical-AI multiple" the June peak implied (~25×+).

The base call as a scoreable claim (logged mentally, not via our model): LGE FY2026 consolidated operating profit ≥ KRW 3.5T, p≈0.55 — i.e., the operating step-up largely happens, but the valuation re-rating that drove June is the fragile part.

Bull vs Bear

Bull case. A century-scale appliance/HVAC manufacturer with real IP (compressors, thermal, motors) is mid-transformation from low-multiple hardware to higher-multiple B2B + platform + recurring: HVAC/data-center cooling is a credible AI-infrastructure adjacency where LGE is an incumbent (not a pretender), spec'd into Nvidia's DSX reference design; subscriptions (~KRW 2.5T, +29%) and webOS ads (>KRW 1T, targeting 20% of OP by 2030) add annuity/margin; VS backlog gives multi-year visibility; and there's embedded SOTP value (LG Display 36.7% + LG Innotek ~30% + a listed India unit worth ~$8.7B at IPO). Management has shown allocation discipline (killed smartphones, IPO'd India). At ~10× forward earnings post-crash, you're paying a hardware multiple for an option on the mix-shift plus a free robotics call.

Bear case (things that could permanently impair or de-rate). (1) TVs (MS) are in structural share loss to Chinese scale players — a segment nearly a quarter of revenue running at an operating loss, with commoditization that brand/OLED can't fully arrest; worst case it's a permanent margin sink. (2) The data-center-cooling thesis is a derivative bet on the AI-capex cycle — a hyperscaler capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. pause guts the one clearly re-rating growth line, and the June→July halving shows how fast that sentiment reverses. (3) Structural low ROE (~4.6%) — even executing the mix-shift, through-cycle returns may stay mediocre, capping the multiple. Pre-mortem (18 months out, thesis broke): the AI/robotics narrative fully deflated, LGE round-tripped to its pre-2026 range (~KRW 120–150k); data-center cooling orders proved lumpy and lower-margin than hoped; TV losses widened; the LG Display equity-method line took an impairment; and robotics remained a cash-absorbing science project with no P&L. Are current multiples too high? After the crash, no — ~10× forward is defensible for the operating business. At the June peak (~392k, ~25×+), yes, plainly. Contrarian view: the market is oscillating between two wrong prices — it over-paid in June for a "robotics company" LGE isn't, and it may now under-appreciate a genuine, sourced HVAC/data-center-cooling + India-SOTP re-rating that doesn't need humanoids to work. The boring bull case is the real one.

Devil's Advocate (short-seller)

Dismantling the bull case:

• What structurally breaks the money-machine: the profit engine is HS appliances at ~7–8% margin in a mature, tariff-exposed, input-cost-sensitive category — there's no growth there, only defense; and the loss engine (MS/TV) is structurally impaired by Chinese scale. Strip out the AI narrative and you own a low-single-digit-margin, low-ROE hardware conglomerate.
• Revenue concentration / what shifts: less single-customer concentration than segment concentration — ~29% of revenue (MS) loses money, and the celebrated new growth (ES/data-center cooling) is a derivative of Nvidia/hyperscaler capex the company doesn't control. If AI-capex growth decelerates, the entire 2026 re-rating rationale evaporates — as the −53%-from-peak move already previewed.
• Weakest-moat point bulls underestimate: in TVs, brand + OLED leadership did not stop LG from falling to ~9% unit share behind TCL and Hisense — proof the moat is porous where Chinese cost structure competes. Extrapolate that risk to any category the Chinese majors target next (appliances included).
• Most dangerous competitor bulls underrate: Samsung Electronics (in premium TV via QD-OLED, and everywhere in the group's shadow) and the Chinese bloc (TCL/Hisense/Midea/Haier) in both TV and appliances — Midea/Haier are formidable, cost-advantaged appliance competitors, not just TV players.
• Worst capital-allocation / incentive flags: the chaebol cross-shareholding web — LGE part-owns two struggling/cyclical affiliates (LG Display, LG Innotek) whose fortunes it can't fully control and whose stakes can be impaired; group priorities can trump minority holders. And the CEO change at the narrative peak removes the strategy's author.
• What must hold for today's price (~185,700): that FY2026 OP genuinely jumps ~50%+ to ~KRW 3.8T and that at least some AI-infra/robotics optionality stays priced in. If OP merely holds near 2025's KRW 2.48T, the forward P/E isn't ~10× — it's ~15–19×, and the stock is not cheap.
• If growth disappoints 20–30%: an OP miss to ~KRW 2.9–3.0T (vs ~3.8T street) likely takes the stock back toward the KRW 130–150k pre-hype zone (−20 to −30% from here) — the "AI premium" is the first thing to go.
• Single scenario that permanently impairs: a durable AI-capex winter coinciding with continued Chinese share-taking in TV and appliances, plus an LG Display impairment — turning LGE back into a sub-3%-margin, sub-5%-ROE hardware maker with no growth narrative. Plausibility: moderate — not a base case, but far from tail.

Management Questions (ordered by information value)

1. TV/MS ran a full-year 2025 operating loss at ~29% of revenue — what is the explicit plan and timeline (exit, JV, restructuring, or fight), and at what point do you stop subsidizing volume share?
2. Data-center cooling: give hard numbers — 2025/2026 order backlog, revenue, and gross margin for DTC/immersion/CDU — and how much is contracted vs. pipeline? How dependent is it on Nvidia specifically?
3. What is the through-cycle ROE/ROIC target, and how does the mix-shift to B2B/platform/subscription move it above the current ~5%?
4. On the CEO transition — what changes and what is deliberately continuous in strategy, capital allocation, and the AI/robotics push under the new leadership?
5. Quantify the Nvidia partnership's economics: is LGE a preferred/qualified thermal vendor with volume commitments, or an MOU? What revenue does it underwrite over 2026–2028?
6. Robotics (Bear + CLoi): what is the current revenue, loss, and path to profitability, and what would make you stop funding it? Is humanoid a real product roadmap or R&D?
7. Subscription/rental (~KRW 2.5T): what's the margin, churn, and financing/receivables profile, and how do you account for the hardware-plus-service bundle?
8. How do you think about the holdco/SOTP discount — with a separately-listed India unit (~$8.7B IPO) plus LG Display/Innotek stakes, would you consider surfacing more value (further listings, stake monetization, buybacks)?
9. Tariffs: quantify the 2025–26 US-tariff P&L impact and the Mexico→Tennessee re-shoring cost/benefit — what's the steady-state US-manufacturing margin?
10. webOS platform: reaffirm or update the "20% of operating profit by 2030" ambition — what's the current run-rate, and what are the risks (ad market, panel-share loss shrinking the installed base)?
11. VS: how durable is the record backlog given EV-demand wobble, and what's the normalized margin ceiling for vehicle components?
12. Capital returns: what is the committed dividend/buyback policy through 2028, and how is it prioritized against robotics/AI investment and group needs?
13. LG Display (36.7%) and LG Innotek (~30%): under what conditions would you impair or exit these equity-method stakes, and how do you insulate LGE minorities from affiliate losses?
14. Where would you deploy the next KRW 1T of growth capex — HVAC/cooling, VS, platform, or robotics — and what hurdle rate?
15. What is the market most wrong about LGE right now — the June over-enthusiasm, or the summer skepticism?

---