brain/
sourcestock-market

Eaton (ETN): COL Group deal, data-center backlog, and Q3 checkpoint (Sep 2026)

Facts behind ETN value after the €810M COL Group deal and the Wells Fargo $503 initiation: Q2 2026 Electrical Americas backlog/orders/DC trend, guidance and Laguna update, COL terms and a computed multiple, the Q3 checkpoint metrics, valuation vs its own history and peers, sell-side targets, and downside evidence.

Source

Eaton (ETN): COL Group deal, data-center backlog, and Q3 checkpoint (Sep 2026)

Generated by Grok Bot research on 2026-09-25. WebSearch + fetch ladder (WebFetch, then curl; SEC EDGAR direct). Native X skipped: spend cap. Treat as raw material. Review before promoting into a project or thread. Facts and labeled computations only. No add/hold/trim call.

Summary

The latest print (Q2 2026, furnished Jul 31) showed Electrical Americas (EA) backlog of $15,175M (+33% YoY), EA book-to-bill of 1.3, and EA rolling-12-month orders +41%. Data-center revenue was up about 65% and data-center orders about 85%. FY26 guidance is organic growth of 11–13%, segment margins of 24.1–24.5%, and adj EPS of $13.40–$13.60. EDGAR shows no 8-K guidance update since then. At Morgan Stanley's Laguna conference in mid-September, the CEO said Eaton is targeting the high end of its Q3 and FY ranges. The COL Group deal is small next to Eaton: €810M EV, €250M forecast 2027 sales (3.2x EV/2027 sales, computed), Q1 2027 close. The release gives no financing or accretion details. The stock is priced above its own history: forward P/E is 29.3x, compared with year-end forward P/Es of 19.5–28.2x for 2021–2025. The Wells $503 target sits near the top of a cluster of $485–$520 targets set since Aug 3. Still open: Q3 date confirmation from Eaton IR (eaton.com was unreachable), segment-level margin guidance, and COL financing/accretion.

Findings

Priors (vault) and what this clip adds

The vault page vault/projects/stock-market/wiki/entities/etn.md (updated 2026-08-27) already carries the Q2 8-K numbers: EA $3,951M / $1,088M OP / 27.5%, FY26 adj EPS $13.40–$13.60, 10-Q firm RPO $24.1B, Electrical backlog +43%, and EA backlog +33%. It also flags the Q1 "228 GW / +240% DC orders" figures as unrestated. This clip adds:

  • the EA backlog in dollars
  • the Q2 call's restated DC figures
  • the Laguna update
  • COL deal terms
  • valuation vs. history and peers
  • dated targets
  • checkpoint metrics

1. Current facts behind value

Electrical Americas backlog/orders (10-Q MD&A, primary).

  • EA backlog was $15,175M at June 30, 2026, vs $11,377M a year earlier (+33%). Organic backlog change was +33%, and organic customer-order change +41% (prior year +2%).
  • Book-to-bill was 1.3 vs 1.1 (Eaton 10-Q, Q2 2026).
  • Electrical Global backlog was $3,602M vs $1,771M (+103%; organic +54%), with book-to-bill 1.1 (same 10-Q).
  • On the call, the CEO said EA backlog has grown $5B since the beginning of 2025, including +$700M sequentially in Q2 (Q2 call transcript, stockanalysis/Quartr).
  • Implied backlog added per year of organic growth: EA backlog grew $3,798M YoY while EA Q2 sales were $3,951M. That is roughly one quarter of sales added to backlog over 12 months (computed from the 10-Q).

Data-center trend.

  • CFO: EA organic growth of 18% was "driven primarily by strength in data centers up about 65%" (Q2 transcript).
  • In Electrical Global, DC organic revenue was also +65%, against an underlying market the CEO put at +23% (Motley Fool Q2 transcript).
  • DC orders about +85%, DC revenue about +65%. Both were repeated by the CEO at Laguna, where he also said the DC negotiation pipeline was up more than 130% (MarketBeat Laguna recap, Sep 17). The Q2 analyst deck reportedly carries the line "Electrical Sector data center orders are up approximately 85%… versus 2Q25". That is a search snippet only; the PDF was not fetched.
  • Restated DC quantification. On the Q2 call, "Total U.S. data center backlog has grown to 307 GW, or 15 years of backlog at 2025 build rates, up from 12 years in our last update. Only roughly 20% of this backlog converts near term. The majority will translate to 2028 and beyond deliveries" (Q2 transcript). At Laguna the CEO cited 342 GW of announced data-center projects vs ~50 GW installed, cautioning that much of it "would unfold over a longer cycle rather than becoming 2027 or 2028 revenue" (MarketBeat Laguna recap).
  • Content per MW: the CEO said $3.4M/MW is "the right number" for modeling (Q2 transcript).

Guidance (latest formal: Q2 release, Jul 31).

  • FY26: organic growth 11–13%, segment margins 24.1–24.5%, GAAP EPS $10.36–$10.56, adj EPS $13.40–$13.60.
  • Q3 26: organic growth 13.5–15.5%, segment margins 24.6–25.0%, adj EPS $3.46–$3.56 (Exhibit 99.1, SEC).
  • On the call: EA organic midpoint raised 200 bps to 15%, Electrical Global up 450 bps to a 12% midpoint, segment-margin ranges reaffirmed, and Boyd FY revenue raised to $1.8B ($1.5B on Eaton's books) (Q2 transcript).
  • No update since: EDGAR shows no 8-K after the Jul 31 Item 2.02 filing. Filings since then are Forms 3/4/144 only (EDGAR submissions feed).
  • At Laguna the CEO said July and August were "very strong" and that Eaton is targeting the high end of existing Q3 and FY ranges. He declined 2027 guidance and called the $31B 2030 electrical revenue target "a floor" (MarketBeat Laguna recap).

Segment margins (Q2 actuals, 10-Q).

SegmentQ2 26 marginQ2 25 margin
EA27.5%29.5%
Electrical Global19.8%20.1%
Aerospace22.8%22.2%
Mobility13.0%12.1%
  • The EA decline was "a 470 basis point decline from higher commodity inflation, partially offset by a 260 basis point increase from higher sales" (10-Q).
  • EA margin bridge from the CFO: Q2→Q3 +250 bps (150 price/cost, 100 output/productivity), Q3→Q4 +200–250 bps, and H2 vs H1 +450–500 bps. He said the 32% 2030 EA margin target still stands (Q2 transcript).
  • Computed from that bridge: implied EA margin is ~30.0% in Q3 and ~32.0–32.5% in Q4. An EA-specific margin guide (reportedly 28.8–29.2% FY in the deck) appears only in a search snippet of the unfetched PDF and is not verified.

2. What COL adds

Terms and footprint. Eaton signed to acquire COL Group from Oaktree's Power Opportunities strategy for an enterprise value of €810M. "COL Group has forecasted sales of €250 million for 2027." Other deal facts from the release (Business Wire release via FinancialContent, Sep 25, 06:45 EDT):

  • Products: medium-voltage distribution, including SF₆-free switchgear, grid automation, and modular power systems.
  • About 400 employees, with facilities in Turin, Milan, Bergamo and Catania.
  • The deal "will expand Eaton's European power distribution capabilities and manufacturing footprint" for data center and utility customers.

PE Hub adds that COL was founded in 1920 near Turin and has four production sites (PE Hub). Only the headline and dek were retrievable.

Price vs sales (computed).

  • €810M / €250M = 3.24x EV / 2027E sales. The denominator is COL's own forecast for a year after close.
  • For reference, Eaton itself trades at 6.38x trailing EV/sales (stockanalysis ETN statistics). The bases differ (forward vs TTM), so treat this as context only.
  • Scale (computed, currency-mismatched): €250M is roughly 1% of Eaton's 2025 revenue of $27.4B (from the release boilerplate). It is roughly 2–3% of Electrical Global's annualized Q2 sales run-rate ($2,517M × 4).

Financing / accretion / conditions.

  • The release does not state financing, accretion/dilution, or EBITDA (release).
  • Close is expected in Q1 2027, "subject to customary closing conditions and regulatory approvals."
  • Balance-sheet context at Jun 30: cash $483M, short-term investments $212M, short-term debt $2,091M, long-term debt $18,509M. Eaton "does not intend to pursue share repurchases in 2026 due to the acquisition of Boyd Thermal." The revolver was upsized to $4.0B, with no borrowings outstanding at Jun 30 (10-Q).
  • At Laguna the CEO said Eaton will keep doing bolt-ons, especially in electrical, but does not expect deals as large as Boyd over the next couple of years (MarketBeat Laguna recap).

EMEA context.

  • On the Q2 call, EMEA revenue was up 20% organically, and the CEO credited EMEA with an 18-month turnaround ("keep expanding margins, gaining share") (Motley Fool Q2 transcript).
  • Electrical Global's organic backlog excluding Boyd was +54% YoY (10-Q). That is the capacity-constraint backdrop COL is meant to address.

3. Next dated checkpoint: Q3 2026 earnings

Date: ~Nov 3, 2026, unconfirmed. Stockanalysis lists "estimated" Tuesday, Nov 3, before market open (stockanalysis ETN statistics). I could not confirm this with Eaton IR: eaton.com timed out on WebFetch and failed on curl this session.

Observable metrics that would confirm or break the DC thesis. Q2 baselines are from the 10-Q, Exhibit 99.1 and the Q2 transcript.

MetricQ2 26 baselineConfirmsBreaks
EA rolling-12M organic orders+41%Holds or acceleratesSharp decel (e.g., into the teens/20s)
EA book-to-bill (rolling 12M)1.3≥1.2≤1.0 (backlog burning)
EA backlog ($)$15,175M (+$700M QoQ)Sequential $ increaseSequential decline
Electrical Sector DC orders YoY~+85%Still strongly positiveFlat or negative
DC revenue YoY~+65%SustainedSharp step-down
EA operating margin27.5%~30% (CFO bridge, computed)Flat vs Q2 / price-cost not neutral
Total segment margin vs Q3 guide24.6–25.0% guideAt/above topBelow range
Q3 adj EPS vs guide$3.46–$3.56 (midpoint $3.51, computed)≥$3.56 (high end targeted)Below range
FY26 organic growth11–13% guideRaised or high endCut
US DC backlog (GW)307 GW (Q2 call); 342 GW "announced" (Laguna)GrowsShrinks or dropped from deck
Boyd FY revenue$1.8B guide ($432M in Q2)Held or raisedCut

4. Priced-in evidence

ETN valuation (stockanalysis/S&P Global, Sep 25, ~10:14 AM EDT, price $440.00) (statistics):

  • Market cap $170.9B; EV $191.5B
  • Trailing P/E 44.8; forward P/E 29.26; EV/EBITDA 28.85
  • Debt/EBITDA 3.07
  • 50-day MA $419.24; 200-day MA $386.86
  • Short interest 1.86% of float
  • Consensus target $479.57 (27 analysts)

Own history, year-end snapshots (stockanalysis ETN ratios):

SnapshotForward P/EEV/EBITDA
FY202123.9022.67
FY202219.4617.93
FY202324.5421.39
FY202428.2424.90
FY202523.8421.31
Current29.2628.85

Both current multiples are above every year-end value shown. Caveat (inference): trailing EBITDA includes Boyd only from Mar 12, 2026, while EV includes its acquisition debt. That likely inflates trailing EV/EBITDA versus a full-year run-rate.

Computed: $441 / $13.50 (FY26 adj EPS guide midpoint) = 32.7x FY26. Against Zacks' FY27 EPS estimate of $15.83, cited on MarketBeat forecast, it is ≈ 27.9x FY27.

Peers (stockanalysis statistics, Sep 25 intraday):

CompanyForward P/EEV/EBITDAConsensus targetPriceSource
Vertiv (VRT)31.9035.34$338.15$251.75VRT
Schneider (EPA:SU)25.8220.16n/an/aSU
Hubbell (HUBB)21.5119.65n/an/aHUBB
GE Vernova (GEV)45.5662.48n/an/aGEV

Takeaway: ETN's forward P/E sits between Schneider/Hubbell and Vertiv/GEV. GEV's trailing EV/EBITDA looks distorted relative to its own P/E; treat it as noisy.

Sell-side targets with dates (MarketBeat forecast table):

DateFirmActionTarget
Sep 24, 2026Wells Fargo (Joseph O'Dea)Initiated Overweight$503
Sep 7, 2026UBSUpgraded to Buy$450 → $515
Aug 28Morgan StanleyOverweight$500 → $520
Aug 20BairdOutperform$500
Aug 3RBCOutperform$484 → $512
Aug 3CitiBuy$471 → $485
Aug 3EvercoreUpgraded to Outperform$502
Aug 3BMOOutperform$477 → $487
May 6BarclaysEqual Weight$392
  • MarketBeat consensus is $456.50 (18 analysts; high $520, low $295).
  • Wells' stated rationale: growth and revision outlook "strong and justifies an upside bias"; "never been a growth cycle like this one in multi-industry." UBS projects 2028 EPS of $19.95, 7% above consensus (Intellectia summary of notes).
  • Computed: Wells $503 is +14.1% vs $441; the stockanalysis consensus is +8.7%; the MarketBeat consensus is +3.5%.
  • 52-week range is $311.92–$478.00 per MarketBeat, so $441 is ~7.7% below the 52-week high (computed).

5. Main downside risks, with evidence

  • DC capex digestion / hyperscaler guidance. The latest round shows no cuts. As of Sep 13, every one of the seven tracked AI builders (Amazon ~$220B 2026; Alphabet $195–205B with a "significant increase" flagged for 2027; Meta $130–145B; Microsoft FY27 growth; Oracle, CoreWeave, Nebius) "raised or held its guidance… None cut" (Supercycle hyperscaler capex roundup, Sep 16). Watch-points from the same source:
    • Alphabet's Q2 was its first negative-FCF quarter since listing.
    • Oracle's capex is partly customer-prepaid.
    • The neoclouds fund capex with debt, "the part of the buildout that can stop faster than it started."
    • Amazon attributed part of its raise to higher memory prices, i.e. dollars that aren't physical build.
    • Next update: late October.
  • Timing risk in Eaton's own DC backlog. Only ~20% of the 307 GW converts near term, with most in 2028+ (Q2 transcript). The CEO cautioned that the 342 GW would not mostly become 2027–28 revenue (Laguna recap).
  • Peer sentiment. Vertiv fell 3.0% on Sep 24 with "no fresh company-specific negative announcement," attributed to rotation out of higher-multiple DC suppliers and higher bond yields (Quiver, Sep 24). VRT trades 34% below its own consensus target (VRT stats).
  • Input costs / tariffs.
    • Q2 gross margin fell from 37.0% to 33.5%, including a "390 basis point decline from higher commodity and wage inflation." EA lost 470 bps to commodity inflation (10-Q).
    • Management expects price/cost to be "roughly neutral" in H2 after the Q2 and early-Q3 price actions (Q2 transcript). April and August price actions plus selective backlog repricing were reiterated at Laguna (recap).
    • The Q2 impact of IEEPA refunds was <$3M ($2.8M), and the tariff-refund impact for H2 is "immaterial" (Motley Fool transcript).
  • Competition / capacity adds.
    • Vertiv agreed on Sep 2 to acquire UtilityInnovation Group to expand AI-DC power capabilities (Quiver).
    • Eaton itself is ramping 24 facilities, 16 of them in ramp-up (Laguna recap).
    • The CEO acknowledged "we know the product lines where our lead times are extended" (Q2 transcript).
  • Leverage / integration load.
    • Q2 net interest expense was $201M vs $71M a year earlier. Boyd ($9.55B, closed Mar 12) is still in its measurement period, and 2026 buybacks are suspended (10-Q).
    • Deals in flight at the same time: the Mobility/Dana RMT (Q1 2027), COL (Q1 2027), and Boyd turning organic in Q2 2027 (Laguna recap).
  • Insider activity. MarketBeat reports sales by executives Heath B. Monesmith and Adam A. Wadecki in August 2026 (MarketBeat forecast). EDGAR shows Form 4s filed Aug 5 through Sep 3 and a Form 144 on Aug 11 (EDGAR submissions feed). Form contents were not read, so transaction sizes are unverified.
  • Customer concentration. No customer-concentration disclosure was found in the fetched 10-Q text. Gap.

Dated checkpoints

DateEventWhat to watchSource
Oct 28, 2026 (estimated)Schneider Electric Q3EMEA/DC read-across for COL's end marketsstockanalysis SU
Late Oct 2026Alphabet, Microsoft, Meta, then Amazon Q32026/2027 capex guidance changes; whether Alphabet's 2027 "significant increase" gets a numberSupercycle
~Nov 3, 2026 (unconfirmed)ETN Q3 2026Metrics in the table in §3stockanalysis ETN
Mid-Nov 2026CoreWeave / NebiusDebt-funded neocloud capex (the fastest-to-stop tranche)Supercycle
Mid-Dec 2026OracleNet capex ceiling ($70B) and prepaymentsSupercycle
Q1 2027COL Group close (regulatory approvals)Any disclosed financing/accretionBusiness Wire via FinancialContent
Q1 2027Mobility/Dana RMT expected close—Laguna recap; vault prior
Early Feb 2027 (inferred, not confirmed)ETN Q4 2026 printFirst formal 2027 guidance (declined at Laguna); implied Q4 EA margin (~32% per CFO bridge, computed)Inferred from the 2026-02-03 prior-year date on nextearningsdate (search snippet only)
Q2 2027Boyd turns organic—Laguna recap
2027 (undated)Possible investor mid-cycle review—Laguna recap

Contradictions and open questions

  • Vault open question now partly answered. The Q1 "228 GW total DC backlog" was superseded on the Q2 call by "Total U.S. data center backlog… 307 GW" (15 years of 2025 build). At Laguna the CEO said "342 GW of announced data-center projects." It is unclear whether 342 GW is the same metric updated or a different definition ("backlog" vs "announced"). Do not flatten the three numbers.
  • DC orders +85% vs Q1's +240%. These are different quarters and possibly different scopes: +85% is described as Electrical Sector, Q2 vs 2Q25. The deceleration from +240% is real on its face, but the base effects are unknown. The Q2 deck PDF was not fetched, so there is no DC order $ figure.
  • "Organic growth could achieve 12%" (Laguna) vs the 11–13% guide. 12% is the Q2 midpoint, not a new raise. "Targeting the high end" implies ~13%, but that is a CEO remark, not a filed update.
  • EA-specific FY margin guide (28.8–29.2%) appears only in a search snippet of the unfetched Q2 deck and is unverified.
  • COL: financing, accretion/dilution, COL EBITDA/margins, and USD price were not disclosed in the fetched release. The ~$923M USD figure seen in search synthesis was not retrieved from a fetched page. No Reuters article was retrieved.
  • Valuation caveat: trailing EV/EBITDA and P/E are distorted by the Boyd closing date and amortization. Forward P/E is the cleaner history comparison.
  • Customer concentration was not found in fetched disclosures.

Provenance

Method: Grok Bot / WebSearch / fetch ladder (WebFetch → curl; SEC EDGAR direct) / native X skipped Generated: 2026-09-25 Rounds: 3 of 3 URLs fetched: 18 successful, 3 failed Fetch notes:

  • SEC 10-Q, Exhibit 99.1 and the EDGAR submissions JSON were retrieved by curl with a declared User-Agent.
  • Stockanalysis peer/ratio pages were retrieved by curl.
  • eaton.com (IR presentations page, news-release listing) timed out on WebFetch and failed to connect on curl. Q3 date and COL release could not be confirmed on Eaton IR.
  • The Business Wire release was read via its FinancialContent syndication (mirror; byline "Eaton via Business Wire").
  • PE Hub returned only headline/dek.
  • Eaton-hosted PDFs (Q2 analyst deck, complete release) were not fetched, per the PDF rule (non-.gov host). Referenced only as search snippets and flagged.
  • No Reuters URL was surfaced by search, so none was fetched.
  • X skipped: spend cap.

Web sources:

X sources:

  • none. X skipped: spend cap (connector returns 403 at monthly cap).

Grokipedia:

  • not used
Referenced by