Modine Manufacturing

Phase A — Understand the business

Lens 1 · Company Overview

Modine designs, engineers and manufactures thermal-management systems — heat transfer at industrial scale. For 109 years (incorporated Wisconsin, June 1916; the founder's "Turbotube" radiator was standard on the Ford Model T) it was a radiator and heat-exchanger company serving vehicles and buildings. Since "the new Modine" reorg announced in late FY2021, management has run the portfolio through an 80/20 discipline (focus resources on the 20% of products/markets that drive 80% of returns), simplified the segment structure, and steered capital into the highest-growth end markets — above all, data-center cooling.

The business today (FY2026) is two reported segments, becoming three:

• Climate Solutions — 65% of FY26 sales. Data-center cooling + commercial/industrial HVAC&R + heat-transfer components. This is the franchise.
• Performance Technologies — 35% of FY26 sales — heat exchangers and cooling for on-highway (commercial vehicle, auto) and heavy-duty/off-highway (ag, construction) plus stationary-power gensets. This is the segment being spun into Gentherm.
• Effective April 1, 2026, Climate Solutions splits into two segments — Data Centers and Commercial HVAC — so FY2027 reports as Data Centers / Commercial HVAC / Performance Technologies.

Product groups, % of net sales:

Data Centers went from a quarter to over a third of the company in one year. Data-center products: chillers (incl. the new 3-MW TurboChill™ 3+MW), dry coolers, precision air handlers, CRAC/CRAH units, fan walls, rear-door heat exchangers, coolant distribution units (CDUs), and immersion solutions — i.e. both air and the high-density liquid-cooling kit the AI buildout needs. Customers span Hyperscale, Colocation, NeoCloud and Edge.

Customers & contract structure:

• Top-10 customers = 49% of FY26 sales (up from 43% FY25, 40% FY24) — concentration is rising.
• One global technology customer = ~11% of FY26 total sales — the first 10%+ customer in years (none in FY25/FY24). This is the Data Centers anchor.
• Named top customers (alphabetical, across both segments): Carrier, Caterpillar, Deere, Stellantis, Trane, Volkswagen, Volvo. The three largest Data-Centers customers are under confidentiality agreements (unnamed).
• Contracts are mostly individual purchase orders (revenue at point-of-shipment). A subset are 3–5-year long-term sales agreements; a "limited number" of highly-specified products recognise revenue over time. Metals cost pass-throughs exist in many long-term contracts but lag 3–12 months and exclude premiums/fabrication.

The signature contract: a long-term capacity agreement with one data-center customer who paid a $165M up-front deposit, against which Modine "expect[s] to sell more than $4 billion of data center cooling products … during calendar years 2027 through 2029". Customer-funded growth capex + a multi-year, multi-billion contracted backlog from a single buyer — the bull case and the concentration risk in one sentence.

Lens 2 · Supply Chain

Upstream inputs → Modine → end customer, named where the filing names them.

• Raw materials: aluminum, copper, steel, stainless steel (nickel) — "several domestic and foreign suppliers," no single-source dependency claimed; metals repriced monthly with raw-material suppliers, quarterly with fabricated-component suppliers. Tariffs added ~$28M to consolidated cost of sales in FY26 (raw-material cost increase incl. tariff impact).
• The chokepoint — purchased components, not metals: "Due to the increasing market demand for data center cooling products, demand for certain key components is currently outpacing supplier capacity. We began experiencing supply shortages in the fourth quarter of fiscal 2026. These component shortages are negatively impacting our production schedules … for the first quarter of fiscal 2027.". Modine is qualifying new vendors to mitigate. This is the single most important operational fact in the filing — the demand is so far ahead of the supply chain that Modine cannot build fast enough, and it dinged Q4 (see Lens 5).
• Manufacturing footprint: 46 manufacturing facilities (Climate Solutions 28 / Performance Technologies 18) across North America, EMEA, Asia + 4 coatings facilities. Operates in Canada, Mexico, US, Brazil, Germany, Hungary, Italy, Netherlands, Serbia, Spain, UK, China, India, South Korea, UAE. Capacity is being added fast in the US and Canada (new leased data-center plants — Canada revenue went $7.9M FY24 → $332.6M FY26; US PP&E nearly doubled to $244.5M).
• Downstream: hyperscalers, colocation operators, NeoCloud/AI-cloud providers (Data Centers); HVAC&R OEMs, wholesalers, contractors (Commercial HVAC); truck/auto/ag/construction OEMs (Performance Technologies).
• End-market demand pull: the broader liquid-cooling market is ~$4.07B (2026) growing to ~$27.65B by 2033, a 31.5% CAGR. Corroborating the wave: Carrier data-center orders +500% Q1 2026; Trane Americas commercial-HVAC applied bookings +120%.

Lens 3 · Competitive Advantages (moats)

Modine's moat is mid-tier and application-specific, not structural. What it has:

• Engineering depth + breadth in thermal — 109 years, state-of-the-art tech centers (Racine, Grenada MS, Allen TX, Leeds, Pocenia, Söderköping, Mezőkövesd, São Paulo), a full-stack cooling catalog from air to immersion. For a hyperscaler standardising on a thermal partner, a vendor that can do chillers + CDUs + rear-door + immersion + controls is stickier than a point-product shop.
• Switching costs via design-in — customers ask Modine to do R&D/design/validation on new projects; that co-engineering "results in stronger customer relationships." Once a cooling architecture is qualified into a data-center design, swapping it mid-program is costly.
• Capacity as a moat right now — in a supply-constrained market, the firm that has US/Canada capacity and a $165M-funded expansion is advantaged versus rivals scrambling for the same components. The >$4B anchor contract is a switching-cost moat the customer itself paid to build.
• 80/20 operating culture — five years of margin self-improvement; a genuine capability, not a slogan (operating margin held ~10.8% even through a brutal data-center-ramp year).

What it lacks: there is no network effect, no IP lock, no scale advantage over the giants. Bargaining power is asymmetric — when your top customer is ~11% of sales and the top-10 are ~half, the hyperscalers need Modine less than Modine needs them, and they know it. The metals pass-throughs lag and exclude premiums, so Modine eats input volatility for 3–12 months. R&D spend actually fell to $29M FY26 (from $35M FY25, $42M FY24) — light for a company claiming technology leadership in a fast-moving cooling race; it leans on acquisitions and co-development instead.

Lens 4 · Segments

Consolidated:

Climate Solutions (65% of sales — the growth engine):

• FY26 growth driver: data-center product sales +$468M; HVAC technologies +$102M (incl. $119M from FY26 acquisitions AbsolutAire / L.B. White / Climate by Design); heat-transfer +$45M.
• Data-center sales grew 73% to $1.1B in FY26 — corroborated by the segment math.
• The catch: Climate Solutions gross margin fell 350bps (28.9% → 25.4%) on "temporary operating inefficiencies … due to the rapid expansion of manufacturing capacity … for data center products" + higher material/tariff costs. Growth is being bought with margin, for now.

Performance Technologies (35% — being spun off):

Declining, lower-margin, cyclical (ag/construction/commercial-vehicle weakness). Management held its operating income flat via cost-out and a strategic exit from low-margin business. This is exactly the segment 80/20 says to shed — and the RMT does.

Geography (by selling-unit location): US $1,676.9M (53%), Canada $332.6M (the data-center capacity build — was $7.9M in FY24), Italy $226.8M, Hungary $213.4M, UK $192.7M, Other $538.7M. US PP&E nearly doubled YoY to $244.5M — the physical footprint of the data-center bet.

---

Phase B — Measure performance

Lens 5 · Earnings Result (Q4 + FY2026, reported 2026-05-26)

The print beat and the stock fell 8.5% — that tension is the whole story.

• Q4 FY26: record quarterly net sales $954.4M, +47% YoY, beat ~$920.7M consensus; adjusted EPS $1.71 vs $1.55 consensus (beat).
• FY26: net sales $3,181M, +23%; data-center sales +73% to ~$1.1B. Record adjusted results — management framed it as a potential 5th consecutive record year.
• GAAP net earnings $123.3M / diluted EPS $2.26, DOWN from $185.5M / $3.42 — entirely because of a one-time $116.1M non-cash pension-termination charge (Modine terminated its primary US pension plan, contributing $15M to fully fund it and buying annuities; the charge recognised actuarial losses parked in AOCI). Strip the pension charge and operating income actually rose $59M to $342M. The GAAP EPS decline is optical; the operating trajectory is up.
• Margin is the real worry: consolidated gross margin −190bps to 23.0%; Climate Solutions −350bps to 25.4% — data-center-ramp inefficiencies + tariffs. The market read "growth at the cost of margin" and, combined with the Q4 component-shortage production disruptions, sold the beat.
• Guidance (FY2027): net sales +20–35%; adjusted EBITDA $650–680M (>40% growth, +100–200bps margin); data-center sales +60–80%. Guidance includes Performance Technologies for the full year (RMT not yet closed). Tone: confident, ramp-continues.
• Balance-sheet flags: trade receivables $731M (+53% YoY) and inventories $506M (+48%) both ran well ahead of 23% revenue growth — a classic ramp working-capital build, partly offset by accounts payable +$151M and contract liabilities +$159M (the customer deposits). Total debt ~$436M, cash $73.5M → net debt ~$363M; net leverage <1x against the 3.5x covenant.
• Market reaction verdict: the tape told you expectations were sky-high. A 47% revenue beat that drops the stock 8.5% is a stock pricing perfection, reacting to the first crack.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on disk (transcripts/ empty); this lens is ``.

• Q4 FY26 call (2026-05-27): record-results framing; emphasis shifted decisively to Data Centers as the identity of the company ("pure-play climate solutions"), the >$4B anchor contract, and FY27 guidance. New, more cautious notes that weren't there a year ago: component supply shortages denting near-term production, and margin normalisation as the ramp settles.
• Tone shift over the year: FY25 → FY26 the language went from "expanding the data-center business" to "we are becoming a pure-play, here is a multi-billion contracted backlog, and we are reorganising the company around it." Management stopped talking about Performance Technologies as a growth story and started talking about it as something to spin out. The recurring phrases are "80/20," "mission-critical thermal," "hyperscale and colocation," and increasingly "NeoCloud."
• The credibility check: they have now delivered five years of margin improvement and a 14.7x stock, so the market extends them benefit of the doubt — which is exactly why the 8.5% drop on a beat matters (the doubt is creeping in on margin and supply).

Lens 7 · Comps

Read: Modine at ~38x forward sits roughly in line with Vertiv (~40s) — i.e. the market is now pricing MOD as a data-center-cooling pure-play, before the RMT has actually made it one, and on a far smaller, more customer-concentrated, lower-margin base than Vertiv's. Versus its own history (5-yr median P/E ~22x) it is ~1.7x its normal multiple. GuruFocus pegs GF Value at $136 vs ~$297 price — flagged 118.6% overvalued. I will not fabricate the EV/EBITDA and peer P/E cells the searches didn't return — but the directional conclusion (priced like a pure-play, isn't yet one) is well-sourced. The comps don't say "cheap." They say "priced for the spin and the ramp to both go perfectly."

Lens 8 · Stock-Price Catalysts (what moves MOD)

5-year total return: $100 → $1,467, ~14.7x (Mar 2021–Mar 2026) vs S&P 500 1.77x. Recent: YTD +122.7%, 1-yr +209.4%. What the tape reacts to:

• The data-center narrative — every data-point that says "AI cooling demand is bigger/sooner" (the >$4B anchor contract, +73% DC sales, raised guidance) has driven the multiple. This is the dominant driver; MOD trades as an AI-infrastructure derivative now, not as an industrial.
• Margin & execution — Q4 FY26: beat on revenue/EPS but −8.5% on margin compression + component-shortage production disruption. The market punishes ramp friction hard at this multiple.
• The Gentherm RMT — the Jan-2026 announcement is a structural re-rating catalyst (pure-play premium) and an overhang (deal risk). Analysts have explicitly flagged it as a near-term share-price risk if it slips.
• Macro / AI-capex sentiment — as an AI-capex proxy, MOD is now hostage to the hyperscaler-capex cycle and any "AI overbuild" scare. Pattern: MOD is a high-beta, narrative-driven AI-cooling name. It moves on demand signals and gets whipped on any execution wobble. It is no longer valued on its own earnings history.

---

Phase C — Judge people & books

Lens 9 · Management

• Neil D. Brinker — President & CEO (since Dec 2020), age 50. Ex-Advanced Energy Industries (President & COO) and Regal/industrial operator. He IS "the new Modine" — the 80/20 transformation, the data-center pivot, the FY26 acquisitions, and the Gentherm RMT are all his. Track record on the tape is exceptional: a 14.7x stock and operating margins lifted through the transformation. Professional manager (not founder), and a good one for this restructuring-into-growth stage.
• Michael B. Lucareli — EVP & CFO (since May 2021), age 57, with Modine since 2010 (prior VP Finance/CFO). Continuity + institutional memory through the re-segmentation.
• Eric S. McGinnis — President, Climate Solutions (since Apr 2022; ex-Regal Beloit). Runs the franchise.
• Jeremy M. Patten — President, Performance Technologies (Sep 2025; ex-ATS) — installed to steward the segment into the spin.
• Capital allocation: disciplined and growth-tilted. No dividend (FY25/FY26) — reinvesting. Buyback authorised $100M (Mar 2025), but essentially untouched — only ~$82M remaining was unused, $0 repurchased in FY26 (only tax-withholding shares). Cash is going to capex ($143M, +$59M, mostly data-center) and M&A ($182M FY26). ROE is healthy (net earnings $123M, depressed by the pension charge, on ~$1.2B equity; on a clean ~$240M+ basis ROE is ~20%+). The RMT brings $210M cash earmarked entirely for debt paydown — clean, deleveraging.
• Red flags: light. Comp plan tied to Adjusted EBITDA growth + EBITDA margin + 3-yr cash-flow ROIC — reasonable. No related-party deals of note. The honest critique: management is buying growth with margin and working capital, and the strategy is increasingly bet-the-company on one end market and (soon) a handful of customers. That's a deliberate, well-executed concentration — but it is concentration.

Lens 10 · Forensic Red Flags

Forensic lens. Label everything.

• Receivables outrunning revenue — the headline flag. Trade receivables +53% to $731M vs sales +23%; 53% of receivables are due from the top-10 customers (vs 43% prior). The cash-flow statement shows receivables drained −$222.6M of operating cash in FY26. Explanation: the data-center ramp + large project shipments late in the year. Why watch it: at rising customer concentration, a single large customer's payment timing or dispute now swings cash materially. Not fraud-shaped — but it's the line a short would lead with.
• Inventory +48% to $506M (drained −$125.1M of OCF) — building ahead of the ramp; reasonable but ties up cash and carries obsolescence risk if DC demand timing slips.
• Earnings quality / the customer-deposit crutch. OCF $248.7M looks healthy, but it is flattered by +$159.0M of contract-liability inflow (the $165M customer deposit) and +$151.1M accounts payable. Net of the contract-liability swing, OCF would be ~$90M against $123M net earnings. FCF = OCF $248.7M − capex $143.3M ≈ $105M — down from ~$129M FY25 despite far higher sales, because capex doubled. The deposit makes FY26 cash-generative; if the next leg of the ramp isn't customer-pre-funded, FCF compresses.
• The $116.1M pension-termination charge — non-cash, ran through AOCI, net $13.1M tax benefit; legitimately one-time, and it removes a long-tail liability (clean-up, not a gimmick). But note it cut GAAP EPS in half — anyone screening on GAAP P/E sees a worse number than the business earned.
• SBC modest — $22.1M (~0.7% of sales); non-GAAP is not being heavily flattered by stock comp. Clean.
• Goodwill $292.1M, entirely Climate Solutions; tested Feb 28 2026, all reporting units' fair value exceeded carrying value, and a 10% FV haircut still wouldn't impair. Accumulated impairment of $40.8M sits historically in Performance Technologies (the segment leaving) — so the franchise's goodwill is clean.
• Other noncurrent liabilities ballooned $108M → $321M — primarily the long-term portion of the customer deposit / contract liabilities. Watch this unwind as the >$4B contract ships (CY27–29).
• Tariffs added ~$28M to cost of sales and are an ongoing margin headwind; OBBBA tax-law changes raised the FY26 provision modestly.

Regulatory findings (required sub-section).

• SEC enforcement: None. regulatory/regulatory-findings.md (EDGAR EFTS, 2021-06-22→2026-06-22): 0 Litigation Releases, 0 AAERs naming Modine.
• 10-K Item 3 (Legal Proceedings): incorporated by reference to Note 20. Note 20 discloses only ordinary-course litigation — "any additional loss in excess of amounts already accrued would not have a material effect." Environmental remediation accruals $12.7M (US facilities + a former Netherlands plant; soil/groundwater from historical operations) — immaterial and declining (was $15.8M). No class actions, no Superfund PRP designations disclosed.
• Non-SEC enforcement (web): no material FTC/DOJ/FDA/EPA actions surfaced for Modine Manufacturing in search.
• Conclusion: No material regulatory or legal findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 / Note 20 as of 2026-06-22. Clean book, clean record. The risk here is operational and valuation, not forensic.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (EPS, next three fiscal years: FY2027–FY2029)

Built bottom-up from FY26 actuals + management's FY27 guidance. Note the structural complication: FY27 guidance includes Performance Technologies, but the RMT removes PT around end-CY2026 (mid-FY2027) — so reported FY28+ EPS is for a smaller, higher-margin RemainCo, and is genuinely hard to model cleanly. Adjusted (non-GAAP) EPS is the right metric; GAAP carries pension/transaction noise. Anchors: FY26 adjusted EPS ≈ $5.50–$5.65; FY27 adjusted EBITDA guide $650–680M.

Method: FY27 adj. EBITDA midpoint ~$665M; subtract D&A ~$85M, interest ~$32M → adj. pre-tax ~$548M; tax ~25% → ~$411M; ÷ ~54M diluted shares ≈ ~$6.75 base adj. EPS. Bull/bear flex EBITDA, margin and the supply-shortage timing. FY28–FY29 are deliberately not point-estimated — the RMT changes the share count (Modine shareholders also receive Gentherm stock, not a Modine buyback), the revenue base (−35%), and the margin mix; modeling a precise post-spin RemainCo EPS without the Form 10 economics would be fabrication. Directionally: RemainCo is smaller-revenue but higher-margin and faster-growing (DC + Commercial HVAC), so RemainCo adj. EPS likely grows off a rebased FY27, with the $210M debt paydown trimming interest.

Brier forecast — NOT logged (per --watchlist rule: skip forecast.ts create in the breadth loop; only log when genuinely committing to a base case). Candidate for a future committed log: "MOD FY27 adjusted EPS ≥ $6.50, p≈0.55, resolves 2027-05-31."

Lens 12 · Bull vs Bear

Bull case. Modine is the smaller, more nimble pure-play on AI data-center cooling about to shed its dead-weight cyclical segment. The franchise grew data-center revenue 73% to $1.1B, has a >$4B contracted, customer-pre-funded backlog (CY27–29) with a hyperscaler, guides FY27 to +20–35% sales / $650–680M adj. EBITDA / +60–80% DC growth, and the end-market (liquid cooling) compounds 31.5% to ~$28B by 2033. The RMT turns it into a clean ~$2B+, ~15%+ operating-margin data-center/HVAC story with $210M of debt paydown — a structurally better business that should command a pure-play multiple. Management has earned trust (14.7x in 5 years, five straight record years, real 80/20 margin gains). Earnings surprise potential: the supply shortage is demand outrunning supply — i.e. the good problem — and resolving it (qualifying vendors) unlocks deferred revenue at improving margin.

Bear case (2–3 ways this permanently impairs or de-rates).

1. The multiple, not the business, is the risk. At ~38x forward / ~133x trailing vs a 22x 5-yr median, MOD prices in flawless execution of both the ramp and the spin. GF Value $136 vs ~$297. A re-rate to even 25x forward on ~$6.75 base EPS is ~$170 — ~40% downside with no change to the operating story. Q4's −8.5%-on-a-beat is the warning shot.
2. Margin may not be "temporary." Climate Solutions GM fell 350bps in the ramp year. If data-center cooling is becoming a more competitive, more commoditised, faster-iterating market (Vertiv, Schneider, Eaton/Boyd, Ecolab/CoolIT all pushing CDUs/immersion), Modine — the sub-scale player with falling R&D — could find the "temporary" inefficiency is structural margin pressure from competition.
3. Concentration + the spin. Post-RMT the company is "less diversified … more vulnerable … concentration of demand among a limited number of large customers or projects" (Modine's own risk factor). One customer is already ~11% and that one contract is >$4B of CY27–29 revenue. If that hyperscaler trims/defers (the contract is not a firm guarantee — "no assurance this customer will purchase the amounts projected"), the thesis cracks. Plus deal risk: IRS ruling, Gentherm shareholder vote, separation friction.

Pre-mortem (18 months out, thesis broke): AI-capex digestion arrived; one hyperscaler paused/renegotiated the capacity deal; the >$4B backlog got pushed right; the supply shortage that was "demand-driven" turned out to mask share loss to Vertiv/Schneider in liquid cooling; margins didn't recover because pricing competition intensified; the RMT closed but RemainCo's growth decelerated to ~15%, and a 38x stock re-rated to 18x. Down 50%+.

Multiples too high? Yes — on every historical and most peer-relative measures, this is priced for perfection.

Contrarian view (what the market refuses to see): the bull consensus treats the >$4B contract as a floor. It's better read as a ceiling proxy — it tells you one customer's demand is real and tells you how concentrated and capex-/working-capital-intensive this growth is. The market is paying a pure-play premium for a not-yet-pure-play that is funding its growth with a customer's deposit and its own balance sheet, in a cooling market the hyperscalers are actively trying to multi-source and commoditise.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull case.

• Revenue concentration is the kill-switch. Top-10 = 49% and climbing; one customer = ~11%; one contract = >$4B over three years; 53% of receivables from the top-10. This isn't a diversified industrial — it's a levered bet on a handful of hyperscaler capex budgets, and management is increasing the concentration via the spin.
• The moat is thin and the R&D is shrinking. $29M R&D (down two years running) against Vertiv/Schneider/Eaton-Boyd — all bigger, all spending, all building liquid-cooling portfolios. Modine's edge is "we have capacity and a catalog," which is a cyclical advantage, not a durable one. When the component shortage eases industry-wide, so does Modine's scarcity premium.
• "Temporary inefficiencies" is the most dangerous phrase in the filing. A 350bps Climate Solutions margin drop, blamed on ramp, in a year of 43% growth — what happens to margin when growth normalises to 20% and competition is fiercer? The bull assumes mean-reversion up; the bear sees the start of competitive margin erosion.
• Cash quality is propped by a deposit. Strip the $159M contract-liability inflow and OCF roughly halves relative to net income; FCF already fell YoY. The growth is being financed — by the customer and the revolver (raised to $550M in Dec 2025).
• Worst capital-allocation tells: none egregious — but $182M of FY26 M&A into HVAC bolt-ons while the core needs capacity, and a buyback authorised but unused, suggests cash is fully spoken for by the ramp.
• Assumptions that must hold for ~$297: DC market keeps compounding 30%+; the hyperscaler honours >$4B; the supply shortage is short; margins re-expand +100–200bps; the RMT closes tax-free and on time; and the multiple stays near 38x. Miss any two and you're looking at a 30–50% de-rate. If FY27 growth disappoints 20–30% (say +18% vs +27% base) and margin disappoints, base EPS slips toward ~$5.50 and a multiple reset to 22–25x = $120–$140 — i.e. roughly GF Value, ~55% below spot.
• The single permanent-impairment scenario: the cooling market commoditises as direct-to-chip/immersion standards converge and the hyperscalers dual/triple-source, turning Modine's mid-tier cooling kit into a price-taking commodity. Plausibility: moderate, and rising with every competitor entry (Eaton/Boyd, Ecolab/CoolIT).

Lens 14 · Management Questions (ordered by information value)

1. The >$4B CY27–29 data-center contract is unconditional in headline but "no assurance" in the risk factor — what are the actual minimum-volume / take-or-pay commitments and termination/penalty terms?
2. Climate Solutions gross margin fell 350bps on "temporary inefficiencies." Quantify temporary vs. structural — what is the steady-state DC cooling gross margin once the ramp normalises, and what proves it's not competitive erosion?
3. Customer concentration is rising (top-10 49%, one at 11%). What is the pipeline of other anchor hyperscaler/NeoCloud contracts, and where does concentration peak post-RMT?
4. The component shortage that hit Q4/Q1 — what % of the FY27 DC ramp is gated by it, when does it clear, and is any of it actually share loss rather than industry supply?
5. Post-RMT RemainCo: give us the pro-forma revenue, adjusted EBITDA margin, and growth algorithm for the standalone Data Centers + Commercial HVAC company.
6. How do you defend against Vertiv/Schneider/Eaton-Boyd in liquid cooling (CDUs, direct-to-chip, immersion) with R&D that has declined three years running?
7. The $165M customer deposit funded this year's growth capex — how is the next leg (CY27–29 capacity) financed, and what is normalised FCF conversion once contract liabilities start unwinding?
8. What is the FY27 capex envelope and the expected peak working-capital drag, given receivables/inventory both grew ~50% this year?
9. On the RMT: latest read on IRS private-letter-ruling timing, Gentherm shareholder vote, and the realistic close date — and the contingency if it slips into 2027?
10. Why authorise a $100M buyback and use ~$0 of it — is the signal "shares are expensive" or "all cash is committed to the ramp"?
11. Tariffs added ~$28M to FY26 cost of sales — what's the FY27 tariff exposure and the lag before contract pass-throughs catch up?
12. Capital-allocation priorities for RemainCo post-$210M debt paydown: organic capacity, M&A, buyback, or dividend initiation?
13. Immersion and direct-to-chip are converging fast — what is Modine's roadmap and design-win position at the GPU-rack level, beyond rack-/row-level cooling?
14. NeoCloud customers are newer and less creditworthy than hyperscalers — how do you underwrite receivable risk as that channel grows?
15. What single internal metric do you watch that would tell you the data-center cooling cycle is rolling over before the revenue does?

---