SPX Technologies, Inc. (NYSE: SPXC) — A Genuine Data-Center-Cooling Inflection, Priced for the Wave Never to Break

Independent equity research. Report date: 2026-07-18. All figures USD unless noted; SPX Technologies’ fiscal year ends December 31. Primary sources: SPX FY2021–FY2025 Forms 10-K, Q1-2026 Form 10-Q (period ended 2026-03-28), the 2026 DEF 14A, and the Q4/FY2025 (2026-02-24) and Q1-2026 (2026-04-30) earnings calls, with public market data and peer comparisons against IDEX, Dover, Nordson, Hubbell and ITT.

⚡ Claude’s Take

This block is the author’s own subjective opinion and general information only — not investment advice. The detailed analysis that follows takes no position and sets no price target.

Verdict: HOLD — a genuinely improving multi-industrial with the realest data-center-cooling inflection in its peer group, but bought at the top of its own range on the group’s lowest absolute returns and its single most concentrated growth driver. Accumulate on weakness in the ~$170–190 zone (≈22–24× the FY26 ~$7.95 adjusted-EPS guide, ≈15–16× forward EV/EBITDA — roughly the $188 at which management itself sold 2.66M new shares in August 2025). At ~$212 you pay ~27× forward adjusted / ~40× trailing GAAP / ~20× trailing EV/EBITDA for an ~11–12% ROIC business whose upside is already the consensus. Not a short. Conviction: medium.

SPX is the member of the diversified-industrial cohort (IDEX, Dover, Nordson, Hubbell, ITT — a group where the recurring read is “great business, richest-ever multiple”) that most deserves the growth multiple on the fundamentals and least deserves it on the price. Deserves it, because unlike Dover (flat revenue five years, +1.6% organic) or IDEX (flat three years), SPX is actually inflecting: FY2025 organic growth was +6.3%, operating margin has climbed 8.9%→15.5% and both segments now earn ~24% segment margins, and the data-center-cooling driver is real and large — ~$200M of 2025 revenue (~9% of sales) guided to grow ~70% to ~$350M in 2026, with a ~$700M capacity build underway and HVAC backlog +38% organic. That is not Dover’s ~$100M rounding-error “AI story”; Marley cooling towers and the OlympusMAX dry/adiabatic system are the thermal-rejection layer of the AI data center. Doesn’t deserve the price, because (1) the entire growth raise is one end market — strip data-center and Ingénia and the rest of HVAC is ~3% organic while Detection & Measurement is ~flat in 2026; (2) returns are the cohort’s lowest — a clean ~11.6% ROIC that clears a ~9–10% WACC by only ~2 points, versus Hubbell’s 16.7% and ITT’s ~15%; (3) half the historical growth was bought at full 3–4× sales / 13–18× EBITDA multiples, so goodwill+intangibles now equal ~85% of equity; and (4) the stock sits at the 93.9th percentile of its own decade on price-to-sales with the factor tape screaming crowded momentum (beta ~1.31, positive Momentum, negative Value) and 3-/6-month risk-adjusted returns already collapsed to ~zero as the +74% 2025 re-rate digests.

The framing is quality-momentum, genuinely inflecting — the good version of the data-center trade (real earnings, not just multiple), but with no margin of safety, a single-end-market dependency, and a live 2022 precedent (the SGS fixed-price cooling-project loss that took operating margin to 4.4%) for how a fixed-price cooling backlog can bite. I want it ~15% cheaper — back toward where management raised equity a year ago — before the risk/reward turns, because there you own a ~$8 forward adjusted-earnings stream at ~23× with a fortress 0.3× levered balance sheet and enormous M&A firepower paying you to wait, rather than a peak multiple on a wave the tape has already celebrated. Flip-bullish if data-center orders sustain into 2027–28 and organic breadth widens beyond the DC/Ingénia core and ROIC steps toward the mid-teens as the new capacity fills — proof the inflection is structural, not cyclical. Flip-bearish if HVAC orders roll over or a fixed-price DC-project charge recurs (the SGS analog) while organic falls back to low-single-digits and ROIC stays stuck ~11% — at which point a ~40× GAAP, beta-1.3 industrial de-rates hard. Tag: “The realest cooling inflection in the group — priced as if it can’t cool off.”

📈 Stock Price Action — Five-Year Event Map

Factual price history — not a recommendation and not a price target. Price moves are FACT; the attributed drivers are INTERPRETATION, cross-referenced to earnings dates, 8-K events, guidance changes, and the news feed.

SPX has been a ~5× off the trough. Over the trailing five years the stock fell to a low of ~$41.90 (2022-04-29), then compounded to an all-time high of ~$246.41 (2026-06-22) before digesting back to ~$211.73 (2026-07-17) — currently ~14% off the high, inside a 52-week range of ~$173.83–$246.41. There is no dividend, so total return equals price return. The move is best understood as roughly half fundamentals (EBITDA ~2×'d) and half re-rating (the multiple ~doubled to a record).

Cycle narrative. (1) SPX rode the 2021 SMID-cap bid before the 2022 rate shock, and — more company-specific — a loss on a large fixed-price evaporative-cooling project at SGS plus a South-Africa DBT impairment cratered operating margin to 4.4% and took the stock to ~$42. (2) As that project was de-risked and supply chains eased, margins snapped back and the stock recovered to ~$66. (3) Through 2023 the operating turnaround became undeniable — margin more than doubled to 12.8%, bolt-ons (ASPEQ $421.5M, TAMCO $125.5M) added scale, and management raised guidance repeatedly — lifting the stock ~54%. (4) 2024 layered on the Ingénia (~$292M) and Kranze/KTS (~$340M) deals and the first articulation of a data-center-cooling opportunity. (5) A brief early-2025 tariff/macro wobble took it back to ~$129. (6) Then the defining leg: from March to October 2025 the stock re-rated +74% as the data-center-cooling order book inflected and Q2/Q3 both beat and raised; management opportunistically issued 2.66M shares at $188 in August. (7) Since the June-2026 all-time high of $246, the stock has digested ~14% — a strong Q1 2026 beat-and-raise (adjusted EPS +22%, FY26 guide lifted to ~$7.95) was absorbed rather than celebrated, the market having already discounted much of the wave at a record valuation. (All price levels FACT; attributed drivers INTERPRETATION.)

1. Executive Summary

SPX Technologies is a ~$2.27B-revenue (FY2025) diversified industrial headquartered in Charlotte, NC, built around two segments: HVAC (~67% of revenue) — engineered evaporative cooling (Marley, Recold, SGS), engineered air movement, and residential/commercial heating (Weil-McLain, Williamson, Qmark, Berko) — and Detection & Measurement (D&M, ~33%) — underground utility locators (Radiodetection, Pearpoint, Schonstedt), pipe inspection/rehab (CUES), transit fare collection (Genfare), obstruction and marine lighting (Flash Technology, Sabik, Sealite), and communication technologies (TCI, Kranze). Under CEO Gene Lowe (since 2015), the company has executed one of the cleaner mid-cap industrial turnarounds of the cycle: revenue has roughly doubled since 2020, operating margin has climbed from 8.9% to 15.5%, EBITDA margin from 11.7% to 21.0%, and both segments now earn ~24% segment margins. The story combines ~6% organic growth, ~8pts of bolt-on M&A (15+ deals since 2018), and continuous “80/20/lean” margin improvement.

The investment tension is not quality — it is price versus the durability and breadth of the inflection. Three facts frame it. First, the growth is real but narrowly sourced: the raised 2026 outlook is overwhelmingly a data-center-cooling story — ~$200M of 2025 revenue guided to ~$350M in 2026 (+70%), with HVAC backlog +38% organic — while the rest of HVAC is ~3% organic and D&M is roughly flat in 2026. Second, returns are good but not elite and only modestly above the cost of capital: a clean ROIC of ~11.6% clears a ~9–10% WACC by ~2 points, the lowest absolute return in a cohort where Hubbell earns 16.7% and ITT ~15%; goodwill and intangibles equal ~85% of equity and adjusted EPS ($6.76) runs ~34% above GAAP diluted EPS ($5.03), a wedge that is partly recurring deal cost for a serial acquirer. Third, the price is at a record on the cleanest yardstick — the 93.9th percentile of SPX’s own decade on price-to-sales, ~20× trailing EV/EBITDA (the top of the diversified-industrial cohort), ~40× trailing GAAP and ~27× the forward adjusted guide — after a +74% 2025 re-rate on the data-center theme. The balance sheet is a genuine asset: after an August-2025 equity raise, net debt is just ~$138M (~0.3× EBITDA), giving large M&A firepower against a pipeline management says it can still buy at ~9× post-synergy EBITDA while data-center peers change hands at 20–30×.

The bull case (secular data-center thermal demand, a long capacity-build runway into 2027–28, disciplined bolt-on compounding, margin runway) and the bear case (a cyclical, capital-attracting HVAC-cooling wave dressed as secular growth; a record multiple; single-end-market concentration; ~11% ROIC barely over WACC; zero insider buying; a fixed-price-project precedent) are both credible — which is precisely why the detailed analysis below takes no position and sets no price target.

2. Business Overview

SPX Technologies supplies highly engineered infrastructure equipment through two reportable segments. The corporate lineage matters: this is the successor to the old SPX Corporation, which spun off SPX FLOW in 2015 and, under Gene Lowe, has since divested away from lumpy, low-return businesses (the large-power Transformer Solutions/GD unit was sold in 2021, the source of the $732.8M discontinued-operations gain that inflated 2021 net income) and acquired toward higher-margin, niche franchises. The company rebranded from SPX Corporation to SPX Technologies in August 2022 to signal the completed reshaping. It employs ~4,700 people, generates ~80% of revenue in the United States, and reports no customer above 10% of sales.

HVAC (~67% of FY2025 revenue; ~$1,518M; 24.5% segment margin). Three sub-franchises: (a) Engineered cooling — Marley cooling towers, Recold fluid coolers/evaporative condensers, SGS package cooling, and Cincinnati Fan / TAMCO / Ingénia engineered air movement and custom air-handling — serving industrial, power-generation and, increasingly, data-center customers; (b) Comfort/residential-commercial heating — Weil-McLain and Williamson-Thermoflo boilers, Patterson-Kelley commercial boilers/water heaters; and © Electric heating — Qmark, Berko, Fahrenheat, INDEECO, and the 2023 ASPEQ acquisition (Brasch, Indeeco). Revenue is a mix of project/engineered-to-order (cooling, air handling — longer cycle, backlog-driven) and replacement/weather-driven (boilers, electric heat — shorter cycle, distribution-sold), with an aftermarket/parts tail whose exact percentage the company does not disclose (open question).

Detection & Measurement (~33%; ~$747M; 23.6% segment margin). A collection of niche, often #1/#2 franchises: Radiodetection / Pearpoint / Schonstedt / Sensors & Software (underground pipe-and-cable locators and ground-penetrating radar, pulled by “call-before-you-dig”/811 damage-prevention regulation); CUES / ULC Robotics (sewer/pipe inspection and rehabilitation robotics); Genfare (public-transit fare collection, including the IoT Vendstar); Flash Technology / Sabik Marine / Sealite / Avlite (FAA-regulated obstruction lighting for towers/wind, and marine navigation aids); TCI and the 2025 Kranze (KTS) acquisition (communication technologies — RF test/countermeasures and defense tactical datalinks/interoperability, now a >$200M “CommTech” platform); and Dielectric (broadcast antennas). D&M is lumpier and more project-driven than its steady margins imply — organic growth was −0.2% in 2024 before rebounding in 2025 — but it is the structurally higher-quality segment (see the Competitive Position section).

How it makes money: design-manufacture-sell of engineered equipment, sold through independent manufacturer reps, third-party distributors, retailers, and direct. Recurring revenue is modest and largely aftermarket/consumable (locator calibration/service, fare-system support, cooling-tower parts) rather than contractual subscription — this is an equipment company, not a razor-blade software model.

Verdict: A cleanly reshaped, two-segment engineered-equipment company with genuine leadership positions — but a heterogeneous one, where a project-driven, cyclically-exposed HVAC-cooling business (two-thirds of revenue) sits alongside a higher-quality niche-monopoly D&M segment (one-third). The two do not deserve the same multiple, and the market is applying its richest one to the whole.

3. Industry Dynamics

SPX operates across a dozen distinct end-markets; the ones that matter for the thesis are cooling/thermal, comfort heating, and the D&M niches.

Engineered evaporative cooling — a consolidated oligopoly now flooded with data-center capital. The global cooling-tower market is a ~$4–5B category compounding high-single-digits, historically split among three scale players — Baltimore Aircoil (BAC, owned by AMETEK), EVAPCO (private), and SPX/Marley — plus regional fabricators. It is a genuine oligopoly with real barriers (thermal engineering, field-erection know-how, brand/spec-in on decades-long installed bases, and the physical/logistics cost of shipping large towers). The data-center overlay is the swing factor: AI/hyperscale thermal loads have pulled a wave of demand into evaporative and hybrid dry/adiabatic cooling (SPX’s OlympusMAX), and management sizes its own data-center revenue at ~$200M (2025)→~$350M (2026). But the same economics are attracting capital — BAC, EVAPCO, JCI, Munters, Vertiv and a raft of liquid-cooling/CDU entrants are all expanding — which is the classic Marathon capital-cycle setup for the abnormal margins to mean-revert once the build-out slows. Critically, liquid cooling does not obsolete the tower (server heat still must be rejected to a cooling tower or dry cooler), but it can migrate value toward CDU vendors and change the mix.

Comfort/residential-commercial and electric heating — mature, fragmented, weather-driven. Boilers (Weil-McLain) and electric heat (Qmark/Berko) are replacement-cycle, GDP-and-weather-driven markets with modest structural growth, many competitors (Burnham/U.S. Boiler, Bosch, Rheem, etc.), and #2–#3 positions for SPX. Structurally unremarkable; a stable cash contributor, not a growth engine.

D&M niches — small, regulated, oligopolistic, infrastructure-backed (the best part). Underground locating is a <$1–2B global niche pulled by mandatory damage-prevention regulation (811/one-call), with Radiodetection a clear leader against Vivax-Metrotech and Vermeer/Subsite. Obstruction lighting is FAA/ICAO-certification-gated. Transit fare collection is a procurement-and-integration-heavy oligopoly (Genfare vs. Cubic/Scheidt & Bachmann). Defense CommTech (TCI/Kranze) carries clearance and program lock-in. These markets are too small to attract large diversified entrants and structurally defensible — the higher-quality half of the portfolio, though individually lumpy.

Verdict: mixed — a genuinely attractive niche-industrial portfolio with one cyclically-hot, capital-attracting core. D&M’s markets are structurally good (small, regulated, oligopolistic). HVAC cooling is a good oligopoly experiencing an abnormally good moment that is, by definition, drawing in competing capital; comfort/electric heating is average. The blended industry backdrop is above-average but not the uniform secular-compounder setting the record multiple implies.

4. Competitive Position

There is no single SPX moat — there are three tiers, and naming them per Greenwald’s taxonomy is the whole exercise.

HVAC engineered cooling — a moderate, partly-cyclical moat. The advantage is a blend of (a) intangibles/brand + spec-in (Marley is specified on installed bases and by engineering firms), (b) economies of scale in a 3-firm oligopoly (thermal-engineering depth, field-erection capability, global service footprint), and © an installed-base aftermarket (parts/service on long-lived towers). The financial test: could a competitor replicate it? Partly — BAC and EVAPCO already have, which is why this is an oligopoly, not a monopoly. And the crucial skeptical point: the current ~24% segment margin is cyclically inflated by the data-center demand/price surge; strip that and HVAC margins likely revert toward the high-teens (their 2021–22 level). Pricing power in a supply-constrained boom is transient, not structural. Verdict: a real but moderate and cyclically-flattered moat.

HVAC comfort/electric heating — a weak-to-absent moat. Mature, fragmented, replacement-driven commodity categories with #2–#3 positions. Defensible cash, not a competitive advantage.

D&M — the strongest moat and the crown jewel. Dominant or top-2 share of markets too small (<$1–2B) to attract scale entrants, reinforced by regulatory/certification spec-in (FAA obstruction lighting, 811 damage-prevention, transit procurement, defense clearance) and consumables/aftermarket + switching costs (a utility standardized on Radiodetection locators, a transit agency on Genfare, a broadcaster on Dielectric does not casually re-qualify). This is a genuine portfolio of small local monopolies in Greenwald’s sense — the same shape as IDEX’s niches. The problem for the thesis: it is only ~⅓ of revenue and grows lumpily (organic −0.2% in 2024). The market’s record multiple is not primarily paying for this segment; it is paying for HVAC cooling.

Direct comparison. Against the diversified-industrial cohort, SPX’s competitive position is mid-tier: its blended ~11–12% ROIC sits below Hubbell (16.7%), ITT (~15%) and AMETEK, roughly with Dover/Nordson/IDEX (~10–11%), and its moat is genuinely bifurcated (elite D&M niches, cyclical HVAC core). The distinguishing feature versus Dover/IDEX is that SPX’s organic engine is actually running — but running on one fuel (data-center cooling).

Verdict: A defensible-but-heterogeneous competitive position — a genuine niche-monopoly D&M crown jewel bolted to a larger, cyclically-flattered HVAC-cooling oligopoly. Real moats exist; they are neither uniform nor wide enough to justify pricing the whole enterprise as a secular compounder. If a “moat” claim can’t be tied to a financial outcome that would deteriorate without it, HVAC-cooling’s abnormal margin is the one most at risk when the cooling wave — and the pricing that comes with it — normalizes.

5. Growth History and Forward Opportunities

History — a real doubling, roughly half bought. Revenue rose from $1,128M (2020) to $2,265M (2025), a ~15% CAGR. The MD&A decomposition is the honest lens: FY2025 +14.2% = organic +6.3% / acquisitions +7.8% / FX +0.1%; FY2024 +13.9% = organic +6.2% / acquisitions +7.7%. So roughly half of the headline growth is bolt-on M&A, and the underlying organic rate is a solid-but-unspectacular ~6%. By segment, HVAC organic was +6.1%/+9.7% (the data-center engine), while D&M organic was +6.3%/−0.2% (project-driven, capable of going negative). Operating margin expansion (8.9%→15.5%) is a blend of genuine 80/20/lean operating leverage, favorable data-center mix, and recovery off the 2022 SGS-loss trough — not a clean “economics improve with scale” story, because the acquired intangible-amortization drag (see Financial Quality) grows alongside.

Forward — one big, real driver and a thinner remainder. The forward case is concentrated:

• Data-center cooling is the engine and it is accelerating. Management sized DC revenue at ~$200M in 2025 (~9% of sales) and, at Q1 2026, raised the 2026 growth guide from +50% to ~+70% (→~$350M) as the Olathe OlympusMAX facility came online early. A ~$700M incremental-capacity build (TAMCO Tennessee, Olathe, Madison AL; ~$550M DC + ~$150M custom air handling) is underway, full-ramp mid-2028, with ~$100M of 2026 capex — throughput-constrained, i.e. demand > supply. HVAC backlog is +38% organic, “primarily driven by data center.” Management flags “attractive runway into '27 and '28.”
• The remainder is thin. Ex-data-center and ex-Ingénia, HVAC organic is ~3%; D&M is roughly flat in 2026 (a ~$20M project pulled into 2025). Total company backlog +42% is genuinely strong, but heavily DC- and defense- (Kranze) weighted.
• M&A optionality is real. With ~0.3× leverage and a $2.025B credit facility, SPX has firepower for continued bolt-ons at ~9× post-synergy EBITDA.

Verdict: high-quality core organic growth (~6%) with a genuinely exciting but narrowly-sourced forward driver. This is better-quality growth than Dover’s or IDEX’s flat organic lines — but its 2026–27 acceleration is overwhelmingly one end market, and the through-cycle organic rate for the blended portfolio is mid-single-digit, which is precisely why the model requires continuous acquisition to deliver double-digit totals. The quality of the growth is above-average; its durability and breadth are the open questions the price does not discount.

6. Financial Quality

Revenue/margins. Covered above: ~$2.27B revenue, gross margin 35.1%→40.5% (2020→2025), operating margin 8.9%→15.5%, EBITDA margin 11.7%→21.0%; both segments ~24% segment margin. 2022 was a genuine trough (operating margin 4.4%) from the SGS fixed-price cooling-project loss plus a South-Africa DBT impairment — a one-time distortion to normalize out, and a cautionary precedent for the current fixed-price DC backlog.

Cash quality is clean. FY2025 cash from operations was $335.6M against net income of $244M (CFO/NI 1.37×), and stock-based compensation is only ~$16.7M (~0.7% of sales) — so, unlike many “adjusted-cash-flow” compounders, there is little SBC-inflated cash here. The catch: capex has tripled to ~$92M (2025) from ~$24M (2023) — capex/sales 1.4%→4.1% — to build the DC-cooling capacity, so equity FCF (~$243M) and forward conversion are structurally lower than the historical asset-light picture implied. This is growth capex, not maintenance, but it lowers near-term free cash.

Quality of earnings — a moderately flattering adjusted number. FY2025 adjusted EPS was $6.76 versus GAAP diluted EPS of $5.03 — a $1.73 (~34%) wedge. The bulk is intangible amortization (~$1.40/sh after-tax; the SG&A intangible-amortization line alone was $87.4M in FY2025, up from $43.9M in FY2023 — the growing accounting cost of the roll-up), plus recurring deal items: acquisition/integration costs $23.8M, Kranze retention-comp amortization $24.2M, and a $1.5M refinancing loss. Because these deal costs recur every year for a serial acquirer, adjusted EPS is moderately flattering — honest owner-earnings sit between the $5.03 GAAP and $6.76 adjusted figures. The market is capitalizing the adjusted number (~27× the $7.95 FY26 guide); on GAAP it is ~40×.

Returns — good, not elite, thin spread over WACC. Stripping the leverage noise (reported ROE of 67% and one widely-used data aggregator’s per-share book figure are both unreliable — that source garbles SPX’s book value at ~$10/sh versus an actual ~$44/sh; use the 10-K), a clean ROIC ≈ NOPAT ~$275M / net invested capital ~$2,375M ≈ 11.6%, only ~1.5–2.5 points above a ~9–10% WACC. Goodwill + intangibles of $1,911.6M ≈ 85% of equity ($2,237.5M) and rising roughly in step — so M&A is accretive but not elite, a thin cushion if the DC cycle cools. ROIC has improved off the 2022 trough (a genuine positive versus the fading-ROIC peers), but its absolute level is the cohort’s lowest.

Balance sheet — a fortress, and a real asset. After the August-2025 raise, FY2025 cash was $364M against $501.6M debt, i.e. net debt of just ~$138M (~0.29× EBITDA); current ratio 2.48×; tangible common equity flipped positive (~$326M / 19.2%) from negative in 2021–2024. The $2.025B credit facility (upsized September 2025) leaves enormous dry powder. This is the cleanest balance sheet in the cohort and underwrites the M&A optionality.

Verdict: high-quality earnings and cash, improving returns — but the economics are good, not elite. Cash conversion is clean, the balance sheet is a genuine competitive asset, and returns are rising off the trough. But ROIC only modestly clears WACC, adjusted EPS is moderately flattering, and the recent capex step-up lowers near-term free cash. Economics are improving with scale — but from a mid-tier base, not toward a Hubbell/AMETEK-class return profile.

7. Capital Allocation

Capital allocation is the SPX equity story — the whole model is buy-niche-franchises, apply 80/20/lean, compound. The record is competent-to-good, with two caveats.

M&A — full multiples, discipline in the integration, not the entry price. SPX has done 15+ deals since 2018. Disclosed prices (10-K Note 4): Cincinnati Fan (2022) $145.2M · TAMCO (2023) $125.5M · ASPEQ (2023) $421.5M · Ingénia (2024) $292.0M · Kranze/KTS (2025, defense CommTech) ~$340M on ~$90M revenue ≈ 3.8× sales · Sigma & Omega (2025) $143.3M · Thermolec (2026) ~$140M on ~$35M revenue ≈ 4× sales · ULC Robotics (2020) ~$135M. Cash M&A spend ran $547M (2023) / $292M (2024) / $445M (2025). Multiples paid are ~3–4× sales (≈13–18× pre-synergy EBITDA) — full, not cheap. Management’s own framing (Q1 2026 call) is candid: “our average valuation over our 18 acquisitions before synergies is 10.5 to 11×… with synergies effectively 9× EBITDA… we’re seeing some data-center companies getting acquired for 20, 25, 30× — we will never be there.” The discipline, then, is in the post-deal 80/20/lean margin lift and the refusal to chase frothy DC-asset multiples, not in bargain entry prices. On the evidence, the deals are accretive (segment margins are up, ROIC has risen off the trough) but the marginal-return cushion is thin.

The 2025 equity raise — disciplined use of a rich currency. In August 2025 SPX issued 2.66M shares at $188 (net ~$551M) near all-time highs, cutting net debt to ~$138M and flipping TCE positive. Issuing an expensively-valued currency to pre-fund an M&A pipeline it can still buy at ~9× is textbook disciplined capital allocation — the opposite of the debt-funded top-tick deal.

Dividends/buybacks — none/minimal, correctly. SPX pays no dividend (confirmed — retained earnings rose exactly by net income), and buybacks are minimal (treasury shares actually fell; net issuance is for comp). For a business reinvesting into 20%±return internal capacity and ~9×-EBITDA bolt-ons, retaining and reinvesting is the right call.

Incentive alignment — good skin, but flawed metrics. CEO Lowe owns ~2.20% (~$217M), insiders ~2.88% as a group — real alignment. But the pay plan has two flaws. First, no return-on-capital metric anywhere — the 2025 annual bonus (paid 163.9% of target) keyed on operating income (50%), “adjusted free cash flow” (25%) and adjusted revenue (25%); LTI is 50% PSUs on 3-year relative TSR (89th percentile → 150% payout). A serial acquirer with no ROIC/ROCE governor can mechanically grow absolute EBITDA and revenue by buying more, regardless of the multiple — the same structural flaw flagged at IDEX, Nordson and Hubbell. Second, a generously-defined cash metric: the bonus “adjusted free cash flow” of $469.0M is roughly 2× actual equity FCF (~$243M) and paid out at 200%.

Verdict: competent-to-good capital allocation — a disciplined operator of a full-multiple roll-up with a fortress balance sheet and real insider alignment, undercut by a pay plan with no return-on-capital brake and a flatteringly-defined cash metric. The model works while the acquired assets keep earning above WACC and the DC wave funds organic reinvestment; the risk is the classic one — growth-by-acquisition rewarded without a return governor, at full prices, into a hot cycle.

8. Changes and Headwinds — Last Two Years

M&A wave. Kranze/KTS ($350M, Dec 2024, closed Q1 2025 — defense CommTech into D&M); Sigma & Omega (Apr 2025, HVAC vertical heat pumps/self-contained units); then a Q1-2026 cluster — Thermolec (~$35M revenue, Canadian electric heat), and Air Enterprises + Rahn (~low-$80Ms revenue, custom air handling/coils) acquired via buying Crawford United and divesting its non-core unit (~$60M cash).

Capital structure. August 2025 equity offering (~$551M net); September 2025 credit facility upsized to $2.025B ($1.5B revolver). Net debt cut to ~$138M; 50.18M shares outstanding post-raise.

Operating momentum. Q4/FY2025 and Q1 2026 both beat and raised. Q1 2026: revenue +17.4%, adjusted EBITDA +23%, adjusted EPS +22% to $1.69; FY2026 adjusted-EPS guide raised +$0.15 to ~$7.95 midpoint (~21% EBITDA growth), net of a $0.05–0.10 Section 232 tariff headwind (mostly Q2, no 2027 impact). HVAC backlog $755M, +38% organic.

Data-center capacity build. The ~$700M incremental-capacity program (Olathe, TAMCO Tennessee, Madison AL) with full ramp mid-2028 — the single biggest strategic bet, and the single biggest concentration risk.

Leadership — a watch item. Two segment presidents are departing within a year: J. Randall Data (Global Ops & Data-Center Solutions, effective March 2026) and John W. Swann III (D&M, January 2027); a new IR head joined from Hertz. Turnover in the two roles most central to the DC build and the crown-jewel segment is worth monitoring.

Headwinds. Section 232 tariffs (a FY26 EPS nick); the cyclical/capital-attracting nature of the DC-cooling surge; fixed-price project risk (the 2022 SGS precedent); D&M lumpiness; and integration risk across a busy M&A cadence.

Verdict: The last two years strengthen the operating thesis (real inflection, disciplined balance-sheet management, accretive deals) while raising the concentration and execution stakes (one end market, a large capacity build, leadership turnover in the two key seats). Net-positive on fundamentals, but the risk profile has narrowed onto data-center cooling.

9. Risk Analysis

Overall: The dominant risks are valuation × cyclicality — a record multiple on a growth driver (DC cooling) that is real but capital-attracting and single-sourced, on a high-beta stock. Catastrophic-loss risk is negligible given the fortress balance sheet; the realistic downside is a sharp de-rating if the DC order book fades or a fixed-price charge recurs.

10. Valuation Discussion (Embedded Expectations)

No price target; no recommendation. This section frames what the price implies.

Where it trades. At ~$211.73 (2026-07-17): market cap ~$9.7B, EV ~$9.7B (net debt ~$138M), trailing EV/EBITDA ~20.4×, EV/Sales ~4.3×, P/E ~40× GAAP / ~27× the FY26 ~$7.95 adjusted guide. On its own decade, its valuation percentile ranks read P/S 93.9th (richest-ever), P/E 65th, P/B 67th, composite 75th — the sales multiple, the cleanest yardstick for a serial acquirer with a growing amortization wedge, is at a record. Forward EV/EBITDA on the ~21%-growth guide (~$575M FY26 EBITDA) is ~16.9× — still full but less extreme than the trailing figure.

Comp context. SPX’s ~20× trailing EV/EBITDA sits at the top of the diversified-industrial cohort (Dover, IDEX, Graco, ITT ~12–19×) — appropriate for the group’s best organic inflection, but demanding for its lowest absolute ROIC (~11–12% vs. Hubbell 16.7%, ITT ~15%). It trades below the pure data-center-cooling names (AAON ~45–50×, nVent’s thermal exposure) — the market is pricing SPX as a top-tier compounder with a data-center kicker, not as a pure DC play.

Embedded expectations. The ~5-year move is ~half fundamentals (EBITDA ~2×'d) and ~half re-rating (the multiple ~doubled). A reverse-DCF at a ~9% WACC implies the current EV embeds roughly ~6.6% perpetual free-cash-flow growth — i.e., the market is underwriting that the ~6% organic + M&A + DC-driven margin story persists in perpetuity. An EBITDA cross-check tells the same story: to clear the cost of equity from here, a buyer needs both sustained low-teens EBITDA growth and a persistent premium multiple. In short, the data-center optionality is already in the price, not a free call.

Scenarios (illustrative, not targets):

• Bear — DC orders normalize post-2027, organic reverts to ~3–4%, a fixed-price charge or capacity-underfill pressures margin; EBITDA growth decelerates and the multiple compresses toward the cohort’s ~14–16× → meaningful downside from a ~20× starting point.
• Base — DC ramp delivers ~$350M (2026) and grows into 2027–28, organic ~5–6%, margins hold ~21%, bolt-ons continue at ~9× → mid-teens EBITDA growth, a full-but-defensible ~16–18× forward multiple → roughly a coupon-plus-modest-appreciation outcome.
• Bull — DC demand proves multi-year-secular with SPX taking share, organic broadens beyond DC, ROIC steps to mid-teens as capacity fills → the premium multiple sustains and earnings compound into it.

What the market is pricing correctly vs. incorrectly: correctly, that SPX has the cohort’s best organic inflection and a genuine (not rounding-error) data-center thermal position with a long capacity runway; arguably incorrectly, that this is a durable secular compounder deserving a record multiple, when two-thirds of revenue is a cyclically-flattered HVAC-cooling oligopoly, returns only modestly clear WACC, half of historical growth was bought, and the entire 2026 raise is one end market.

11. Variant Perception

Consensus. SPX is a high-quality mid-cap industrial compounder riding a durable data-center-cooling secular wave, with a disciplined bolt-on M&A engine, a fortress balance sheet and a long margin/capacity runway — worth a premium multiple (Truist PT $295, Jul 2026; the factor tape shows a crowded momentum/industrial-renaissance trade, beta ~1.31, positive Momentum, negative Value, positive Infrastructure/SmallSize).

Strongest bull case. The data-center thermal opportunity is real, large and accelerating (DC revenue ~$200M→~$350M, +38% organic HVAC backlog, ~$700M capacity to mid-2028, runway into 2027–28); Marley/OlympusMAX are the actual heat-rejection layer of the AI build-out; SPX buys at ~9× post-synergy while DC peers change hands at 20–30×; ~0.3× leverage gives huge M&A firepower; both segments’ margins are still expanding; and unlike its flat-organic peers, SPX’s compounding is earned, not just bought.

Strongest bear case. A cyclical, capital-attracting HVAC-cooling wave dressed as secular growth; the richest-ever multiple (94th-pctile P/S, ~40× GAAP) on the cohort’s lowest absolute ROIC (~11–12%, ~2pts over WACC); half of historical growth bought at full 3–4× sales multiples with goodwill+intangibles ~85% of equity; the entire 2026 growth raise sourced from one end market while the rest of HVAC is ~3% organic and D&M is flat; a moderately flattering adjusted EPS; a pay plan with no ROIC governor and a 2×-overstated cash metric; zero insider open-market buying in 181 Form 4s over five years; a high beta and a fixed-price-project precedent (2022 SGS).

The 3–5 assumptions that matter most:

1. Data-center-cooling demand is multi-year-secular, not a 2025–27 build-out spike. (Bull requires; bear disputes.)
2. The ~$700M capacity fills at target margins without a fixed-price charge or underutilization.
3. Organic growth broadens beyond DC/Ingénia so the model isn’t single-sourced.
4. ROIC steps toward the mid-teens as capacity leverages, rather than stalling near WACC.
5. The premium multiple persists — the incremental return now depends on the multiple not compressing, since the easy re-rate is done.

Falsification tests. Bull falsifies if HVAC/DC orders roll over, a fixed-price DC-project charge recurs (the SGS analog), or organic ex-M&A decelerates to low-single-digits while ROIC stays ~11% — the record multiple then de-rates. Bear falsifies if organic (ex-M&A) growth sustains double-digits into 2027–28 and ROIC steps into the mid-teens as the capacity fills — proving the inflection is structural and the premium earned.

12. Fact vs. Interpretation

13. Open Questions

1. Aftermarket/recurring-revenue mix — undisclosed; how much of HVAC and D&M revenue is parts/service/consumable vs. project/OEM? Material to moat durability.
2. Durability of data-center demand beyond the 2026–28 build-out — is this a secular re-rating of thermal loads or a capacity spike that mean-reverts?
3. Fixed-price exposure in the DC backlog — how much of the +38% HVAC backlog is fixed-price, and what is the SGS-style loss risk?
4. Incremental/acquired ROIC — are recent deals (Kranze, ASPEQ, Ingénia) earning above WACC on a standalone basis, or is aggregate ROIC being propped by the organic DC surge?
5. Segment-president turnover — does the departure of the DC-Solutions and D&M presidents signal anything beyond routine succession?
6. Normalized margin — where does HVAC margin settle once the DC-pricing surge normalizes and the ~$700M capacity is fully absorbed?

14. What Must Be True

Bull case — what must be true:

• Data-center-cooling demand is a multi-year secular phenomenon (into 2028+), not a 2025–27 build-out spike, and SPX holds/gains share against BAC/EVAPCO/Munters/Vertiv.
• The ~$700M capacity build fills at target margins with no material fixed-price charge or under-utilization.
• Organic growth broadens beyond DC/Ingénia and ROIC steps toward the mid-teens as capacity leverages, justifying the premium multiple.
• Falsification test: four+ quarters of HVAC/DC order deceleration toward low-single-digit organic, a fixed-price DC-project charge, or ROIC stuck ~11% — any one breaks the secular-compounder thesis.

Bear case — what must be true:

• The DC-cooling surge is a cyclical, capital-attracting spike that normalizes post-2027, dragging HVAC margins back toward the high-teens and organic toward ~3–4%.
• The record multiple compresses toward the cohort’s ~14–16× as growth decelerates and the market re-recognizes an ~11%-ROIC, half-bought-growth industrial.
• Falsification test: organic (ex-M&A) growth sustains double-digits into 2027–28 and ROIC steps into the mid-teens — proving the inflection structural and the premium earned; the bear is then wrong.

15. Source Appendix

See Appendix B (Source Appendix) below for the full list of primary and secondary sources. Primary sources: SPX Technologies FY2021–FY2025 Forms 10-K; Q1-2026 Form 10-Q (period 2026-03-28); 2026 DEF 14A; Q4/FY2025 (2026-02-24) and Q1-2026 (2026-04-30) earnings-call transcripts; SEC Form 4 corpus (2021–2026); Aug-2025 equity-offering and Sep-2025 credit-facility 8-Ks. Public market data (prices, valuation multiples, factor exposures) and peer comparisons against IDEX, Dover, Nordson, Hubbell and ITT.

APPENDIX A — Standard Diligence Questionnaire — SPX Technologies, Inc. (NYSE: SPXC)

Report date 2026-07-18. Supplemental to the research memo. Fact / Interpretation / Assumption labels applied where material.

General

What thoughtful questions have other investors asked about this company? The core debate is whether the data-center-cooling inflection is secular or cyclical, and whether ~11–12% ROIC and half-bought growth justify a record ~20× EV/EBITDA / ~40× GAAP multiple. Sharp questions we hear: (1) How much of HVAC’s ~24% segment margin is a transient data-center pricing surge that reverts when the build-out slows? (2) With the whole 2026 growth raise sourced from data centers, what is the through-cycle organic rate of the rest of the portfolio? (Answer: ~3% ex-DC/Ingénia HVAC, ~flat D&M.) (3) Are the 3–4×-sales acquisitions earning above WACC standalone, or is aggregate ROIC propped by the organic surge? (4) Why has there been zero insider open-market buying in five years if management is as confident as the guide implies? (5) Is the ~$700M capacity build a share-gain masterstroke or a fixed-cost trap if AI-cooling demand normalizes?

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? Interpretation: Above mid-cycle and rising. Operating margin (15.5%) and both segment margins (~24%) are at records, cyclically flattered by the data-center demand/pricing surge; the honest normalized margin is somewhat lower. 2022 (4.4% op margin) was the recent trough.

Driven by the external environment or internal actions? Both — genuine internal 80/20/lean execution and portfolio reshaping (Transformer Solutions divested 2021) plus an external data-center thermal-demand wave. The internal piece is durable; the external piece is the question mark.

How stable are revenues? Moderately. HVAC cooling and D&M are project/backlog-driven and lumpy (D&M organic was −0.2% in 2024); comfort/electric heating is weather/replacement-driven. Recurring/contractual revenue is modest (aftermarket/consumable, not subscription).

Outlook for products/services? Strong near-term on data-center cooling (backlog +38% organic, ~$700M capacity to mid-2028); mid-single-digit for the balance.

How big will this market be? Cooling towers ~$4–5B growing high-single-digit, with a data-center overlay pulling a multi-year build-out; D&M niches are small (<$1–2B each), regulated, GDP-plus. Global, ~80% US revenue.

Business Quality & Competitive Moat

Is the industry getting more or less competitive? More, in the hot part — data-center cooling economics are attracting capital (BAC, EVAPCO, JCI, Munters, Vertiv, liquid-cooling/CDU entrants). D&M niches remain stably oligopolistic.

How profitable is the business (ROIC, ROE)? Fact/Interpretation: Clean ROIC ~11.6% (only ~2pts over a ~9–10% WACC); reported ROE (67%) and some aggregators’ book-value figures are unreliable (thin/leveraged book; one source garbles book at ~$10/sh vs ~$44 actual — use the 10-K). Good, not elite; the cohort’s lowest absolute ROIC.

How profitable is the industry — competitors, barriers? HVAC cooling: 3-firm oligopoly (BAC, EVAPCO, SPX) with real but replicable barriers. D&M: small niches with regulatory/certification barriers — the higher-return part. Comfort/electric heat: fragmented, commodity.

Can the business be easily understood? Yes — engineered-equipment manufacturer in two segments; the complexity is the acquisition accounting (goodwill/intangibles ~85% of equity) and the adjusted-vs-GAAP wedge.

Undermined by foreign low-cost labor? Low risk — engineered, often field-erected/spec’d, freight-sensitive (large towers), and increasingly US-produced (Olathe, TAMCO Tennessee, Madison AL). Tariffs (Section 232) are a modest headwind, not a labor-arbitrage threat.

Do brands matter? Yes, as spec-in intangibles — Marley (cooling), Radiodetection (locating), Weil-McLain (boilers), Genfare (transit) carry real specification and installed-base value, though they are engineering brands, not consumer brands.

Nature of competition? Engineering, spec-in, installed-base/aftermarket, lead-time/capacity (currently demand > supply in cooling), and regulatory certification in D&M.

Customers’ switching costs? Moderate-to-high in D&M (re-qualification, standardization); moderate in engineered cooling (spec-in, aftermarket); low in comfort/electric heating.

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? The internally-built brand/spec-in intangibles and the aftermarket installed base are under-recognized; conversely, acquired goodwill/intangibles ($1.9B) are fully on the books.

Off-balance-sheet liabilities? Nothing unusual flagged — operating leases capitalized under ASC 842; pension is modest/frozen. Acquisition earnouts (e.g., Thermolec) are contingent.

How conservative is the accounting? Interpretation: Reasonable but with a moderately flattering adjusted-EPS presentation (recurring deal costs and intangible amortization added back), and a bonus “adjusted FCF” metric (~$469M) ~2× actual equity FCF (~$243M). Cash conversion itself (CFO/NI 1.37×) is clean.

How CapEx-hungry? Historically light (~1.4% of sales) but rising sharply — capex tripled to ~$92M (2025, ~4.1% of sales) for the data-center capacity build (~$100M in 2026). Growth capex, but it lowers near-term free cash.

Capital Allocation & Management

How much FCF, and how used? ~$243M equity FCF (FY2025), depressed by growth capex. Used for M&A (the priority), debt reduction, and comp-related issuance — no dividend, minimal buybacks.

Significant acquisitions recently? Yes — Kranze/KTS ($350M, 2024–25), ASPEQ ($421.5M, 2023), Ingénia ($292M, 2024), Sigma & Omega, Thermolec, Air Enterprises/Rahn (2026). ~3–4× sales / ~13–18× pre-synergy (~9× post) EBITDA.

Buying back shares? No, minimal. Issuing shares? Yes — 2.66M at $188 (Aug 2025, net ~$551M) to pre-fund M&A and de-lever; plus routine comp issuance. Disciplined use of a rich currency.

Compensation policy / motivations? CEO Lowe comp ~$9.34M; insider ownership ~2.2% CEO / ~2.9% group (real alignment). Flaw: no ROIC/return-on-capital metric in any incentive plan (bonus on op income/adjusted FCF/revenue; LTI on relative TSR + net income) — a serial acquirer with no return governor.

Valuation & Market Data

ADR, MLP, or K-1? No — ordinary NYSE common stock, US C-corp.

Dividend policy? None. All capital reinvested.

How profitable? Operating margin 15.5%, EBITDA margin 21.0%, net margin 10.8%; ROIC ~11.6%.

Net income vs cash from operations diverging? No adverse divergence — CFO ($335.6M) exceeds net income ($244M), CFO/NI 1.37×. The divergence to watch is free cash (equity FCF ~$243M) being pressured by the growth-capex step-up.

Risks & Downside

What would cause the stock to decline? A data-center order rollover; a fixed-price cooling-project charge (2022 SGS analog); a multiple de-rating on a high-beta (~1.31) name from the 94th-pctile P/S; ROIC stalling near WACC; capacity under-fill; broader industrial/tariff shock.

Catastrophic loss risk? Low — diversified, profitable, 0.3× levered, positive TCE, IG-quality credit.

Chance of total loss? Negligible.

Recent News & Events

Has the business environment changed recently? Yes, favorably near-term — a data-center-cooling demand acceleration (2026 DC guide raised from +50% to +70%), a busy M&A cadence, and a de-levering equity raise; offset by a Section 232 tariff nick and segment-leadership departures.

Significant acquisitions? Yes (above). Change in accounting policies? None material. Recent changes — new markets, facilities, management? ~$700M new US cooling/air-handling capacity (Olathe, TAMCO Tennessee, Madison AL); two segment presidents departing (DC-Solutions eff. Mar 2026; D&M eff. Jan 2027); new IR head; 50.18M shares outstanding post-raise.

APPENDIX B — Source Appendix — SPX Technologies, Inc. (NYSE: SPXC)

Report date 2026-07-18. Primary sources over secondary; recent over stale. Access dates 2026-07-18 unless noted.

Primary — SEC filings (SPX Technologies, CIK 0000088205)

Primary — Earnings-call transcripts

Public market data (reconciled to filings)

Note: reported ROE and some aggregators’ per-share book-value figures are distorted by SPX’s thin/leveraged book (one source shows ~$10/sh vs an actual ~$44/sh); the 10-K is authoritative.

Secondary / industry context

All non-obvious facts in this article are drawn from the primary and public sources listed above.