Eaton Corporation plc — Q2 2026 earnings review

Verdict: ACCELERATING on demand, DECELERATING on conversion. Eaton delivered its cleanest double beat in three years and raised the full-year organic guide by 200 bps — four times the size of last year's Q2 raise — while leaving the segment-margin range completely untouched. That is exactly what a management team does when it believes the volume but not yet the drop-through, and it is the right way to hold this quarter.

The demand side is the strongest it has been in a decade. Revenue growth has accelerated in five of the last six quarters to a cycle-high +21.4%, and the acceleration itself is widening (+293 → +378 → +454 bps). Organic alone went from +6% to +14% (+16% ex-Mobility), so this is not an acquisition mirage. Every order and backlog line is at or near an 8-quarter high and each is growing faster than the revenue it feeds: Electrical Global orders went from −1% to +33% R12M in four quarters with backlog +103% YoY (+54% organic ex-Boyd); Electrical Americas orders +41% R12M with backlog +33%; Aerospace backlog +28%; book-to-bill 1.2 in both Global and Aerospace.

The conversion side is the problem, and it is entirely self-inflicted. Gross margin compressed year-over-year for a fifth straight quarter and the compression is widening (−98, −24, −209, −278, −349 bps) to a 12-quarter low of 33.5%. Adjusted EPS grew only +6.8% against revenue +21.4% — a 1,460 bp wedge, the widest in the window. GAAP EPS is outright down 15.9% in a record quarter, and the adjusted-to-GAAP wedge more than doubled year-over-year to $1.04 from $0.44. Interest expense alone (+$130M YoY, roughly −$0.27/share after tax) explains most of the gap. This is the price of the $9.55B Boyd Thermal deal — known, disclosed, self-inflicted, and by management's own bridge, temporary.

Tone: defensive to offensive. Q1 was an explanation of a margin shortfall ("Q1 will be the trough," guide cut 50 bps). Q2 is a claim of momentum — the strategy is "gathering pace," orders accelerating across all electrical end markets, and management still calls itself "very prudent" after a cumulative 450 bps of organic raises. They promised the Electrical Americas sequential recovery on the Q1 call and delivered it one quarter later (+190 bps QoQ to 27.5%), then upgraded the disclosure from narrative to a quantified bridge. That is the single most management-quality-positive item in the quarter.

Contradictions: 10 found, 5 high-materiality, and they cluster in exactly the two places the bull case lives — the Boyd/data-center TAM math and the Electrical Americas ramp and price/cost story. Management is reliable on quantities already banked (the 44/56 EPS split landed at 44.1%) and unreliable on the forward margin and TAM narrative. That is precisely the layer the ~31x FY26E multiple is paying for.

The binary catalyst is the Q3 print (2026-11-03). Q3 adjusted EPS is guided essentially in line with consensus ($3.51 vs $3.499), which loads the entire FY beat into an implied ~26.5% Q4 segment margin — a company record. Underwrite the volume; do not yet underwrite the drop-through.

Consolidated, 12 quarters

| Quarter | Revenue $M | YoY | Gross margin | GM YoY bps | EBITDA $M | EBITDA margin | Adj EPS | Adj EPS YoY | GAAP EPS | |---|---:|---:|---:|---:|---:|---:|---:|---:|---:| | Q2'26 | 8,531 | +21.4% | 33.5% | −349 | 1,791 | 21.0% | $3.15 | +6.8% | $2.11 | | Q1'26 | 7,451 | +16.8% | 35.6% | −278 | 1,444 | 19.4% | $2.81 | +3.3% | $2.22 | | Q4'25 | 7,055 | +13.1% | 36.8% | −209 | 1,632 | 23.1% | $3.33 | +17.7% | $2.91 | | Q3'25 | 6,988 | +10.1% | 38.3% | −24 | 1,625 | 23.3% | $3.07 | +8.1% | $2.59 | | Q2'25 | 7,028 | +10.7% | 37.0% | −98 | 1,517 | 21.6% | $2.95 | +8.1% | $2.51 | | Q1'25 | 6,377 | +7.3% | 38.4% | +107 | 1,461 | 22.9% | $2.72 | +13.3% | $2.45 | | Q4'24 | 6,240 | +4.6% | 38.9% | +145 | 1,541 | 24.7% | $2.83 | +11.0% | $2.45 | | Q3'24 | 6,345 | +7.9% | 38.5% | +117 | 1,494 | 23.5% | $2.84 | +15.0% | $2.53 | | Q2'24 | 6,350 | — | 37.9% | — | 1,444 | 22.7% | $2.73 | — | $2.48 | | Q1'24 | 5,943 | — | 37.3% | — | 1,308 | 22.0% | $2.40 | — | $2.04 | | Q4'23 | 5,967 | — | 37.5% | — | 1,309 | 21.9% | $2.55 | — | $2.35 | | Q3'23 | 5,880 | — | 37.3% | — | 1,314 | 22.3% | $2.47 | — | $2.22 |

Three things jump off this table. (1) Revenue growth has accelerated in five of the last six quarters — +4.6% → +7.3% → +10.7% → +10.1% (the one step back) → +13.1% → +16.8% → +21.4% — the fastest print in the window by 460 bps, with acquisitions contributing 7 of the 21 points and organic the larger half. (2) Gross margin has compressed year-over-year for five straight quarters and the compression is widening; at 33.5% this is 540 bps below the Q4'24 peak. (3) Adjusted EPS has decoupled from revenue — a 1,460 bp wedge, the widest in 12 quarters.

The EBITDA margin line (−59 bps YoY versus −349 bps at the gross line) is the constructive counterpoint: most of the gross-margin damage is recovered below gross profit through operating leverage on SG&A. Revenue is scaling faster than opex.

Segment revenue and margin

| Driver ($M) | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26 | |---|---:|---:|---:|---:|---:|---:|---:|---:| | Electrical Americas | 2,963 | 2,905 | 3,010 | 3,350 | 3,410 | 3,506 | 3,600 | 3,951 | |   YoY | +14.2% | +8.7% | +11.9% | +16.4% | +15.1% | +20.7% | +19.6% | +17.9% | | Electrical Global | 1,573 | 1,569 | 1,610 | 1,753 | 1,724 | 1,728 | 1,945 | 2,517 | |   YoY | +4.7% | +3.8% | +7.3% | +9.2% | +9.6% | +10.1% | +20.8% | +43.6% | | Aerospace | 946 | 971 | 979 | 1,080 | 1,079 | 1,111 | 1,139 | 1,222 | |   YoY | +9.1% | +8.5% | +12.4% | +13.1% | +14.1% | +14.4% | +16.3% | +13.1% | | Mobility | 863 | 794 | 779 | 845 | 775 | 711 | 766 | 841 | |   YoY | −5.8% | −10.5% | −11.7% | −7.3% | −10.2% | −10.5% | −1.7% | −0.5% | | Total segment op profit | 1,544 | 1,542 | 1,522 | 1,682 | 1,749 | 1,760 | 1,690 | 1,976 | |   YoY | — | — | +11.0% | +12.0% | +13.3% | +14.1% | +11.0% | +17.5% |

| Segment operating margin | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26 | |---|---:|---:|---:|---:|---:|---:|---:|---:| | Total segment | 24.3% | 24.7% | 23.9% | 23.9% | 25.0% | 24.9% | 22.7% | 23.1% (−80 bps YoY) | | Electrical Americas | 30.1% | 31.6% | 30.0% | 29.5% | 30.3% | 29.8% | 25.6% | 27.5% (−196 bps YoY) | | Electrical Global | 18.7% | 17.7% | 18.6% | 20.1% | 19.1% | 19.7% | 19.2% | 19.8% (−34 bps YoY) | | Aerospace | 24.3% | 22.9% | 23.1% | 22.2% | 25.9% | 24.1% | 26.7% | 22.8% (+58 bps YoY) |

Orders, backlog and book-to-bill — the leading indicator

| Indicator (YoY) | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26 | |---|---:|---:|---:|---:|---:|---:|---:|---:| | Electrical Americas orders, R12M | +16% | +16% | −4% | +2% | +7% | +16% | +42% | +41% | | Electrical Americas backlog | +26% | +29% | +6% | +17% | +20% | +31% | +44% | +33% | | Electrical Global orders, R12M | +6% | +4% | 0% | −1% | +2% | +6% | +13% | +33% | | Electrical Global backlog | +19% | +16% | +5% | +1% | +7% | +19% | n/d | +103% (+54% organic) | | Aerospace orders, R12M | +6% | +10% | +14% | +10% | +11% | +11% | +13% | +17% | | Aerospace backlog | +14% | +16% | +16% | +16% | +15% | +16% | record | +28% | | Electrical book-to-bill, R12M | 1.1 | 1.1 | above 1.0 | 1.1 | 1.1 | 1.1 | 1.2 | 1.2 |

Quarterly revenue, 12 quarters ($B)

Demand versus delivery — orders lead revenue, revenue leads EPS

Five-year annual trend with forward estimates

| Year | Revenue $M | YoY | Gross margin | EBITDA $M | EBITDA margin | Adj EPS | Adj EPS YoY | GAAP EPS | |---|---:|---:|---:|---:|---:|---:|---:|---:| | 2021 | 19,628 | +9.9% | 32.2% | 3,792 | 19.3% | $6.63 | +56.0% | $5.34 | | 2022 | 20,752 | +5.7% | 33.3% | 4,195 | 20.2% | $7.57 | +14.2% | $6.14 | | 2023 | 23,196 | +11.8% | 36.4% | 4,921 | 21.2% | $9.11 | +20.3% | $8.02 | | 2024 | 24,878 | +7.3% | 38.2% | 5,787 | 23.3% | $10.80 | +18.6% | $9.50 | | 2025 | 27,448 | +10.3% | 37.6% | 6,235 | 22.7% | $12.07 | +11.8% | $10.46 | | 2026E (co. guide) | ~32,100 | +17.0% | — | ~6,740 | ~21.0% | $13.40-13.60 | +11.8% at mid | $10.36-10.56 | | 2027E (street) | 35,550 | +10.6% | — | 7,460 | 21.0% | $15.73 | +16.5% | — |

The five-year picture is the same shape one derivative slower: revenue CAGR of +8.7% against adjusted EPS CAGR of +16.2% — margin expansion did the heavy lifting through 2024 (gross margin +600 bps from 2020 to 2024), then reversed in 2025 (−59 bps) and sharply further in 1H'26. The mix shift is structural: Electrical Americas grew from 40% of segment revenue in 2021 to 48% in 2025 while Mobility fell from 16% to 11% and is now being separated entirely.

Beat rates and magnitude

| Window | Adj EPS beat rate | Revenue beat rate | Avg EPS beat | Avg revenue beat | |---|---:|---:|---:|---:| | Last 4 quarters | 4 / 4 = 100% | 2 / 4 = 50% | +1.70% | +1.81% | | Prior 4 quarters (Q3'24-Q2'25) | 4 / 4 | 2 / 4 | +0.71% | +0.49% | | Four before that (Q3'23-Q2'24) | 4 / 4 | 3 / 4 | +4.55% | +0.29% | | Last 12 quarters | 12 / 12 = 100% | 7 / 12 = 58% | +2.32% | +0.86% |

Pattern: consistent beater on EPS, mixed on revenue — and the magnitude is inflecting on the more credible line. The L12Q EPS record on its own is close to information-free: through 2024-25 the beats were $0.01-$0.02, i.e. rounding. The signal is in the revenue series, which management cannot manage the way it manages an EPS guide. Revenue beat magnitude has gone +0.29% → +0.49% → +1.81%, with the two largest revenue beats of the entire 12-quarter window landing in the last two quarters (+4.30%, +4.61%). In the 2024-25 trough ETN was missing revenue in four of eight quarters while still eking out penny EPS beats — the classic signature of a company managing to a guide against a soft top line. That has flipped.

Quality of the beat — the one caveat

| Quarter | Revenue beat | EPS beat | Drop-through ratio | |---|---:|---:|---:| | Q2 2026 | +4.61% | +2.61% | 0.57x | | Q1 2026 | +4.30% | +2.93% | 0.68x |

The revenue beat did not flow through proportionally, for two acquisition-related reasons that both belong in the thesis. Mix: a meaningful slice of the upside was Boyd Thermal ($432M of Q2 revenue), and Boyd sits inside Electrical Global at a 19.8% margin versus the 23.1% corporate segment margin — structurally dilutive. Below the line: the adjusted-to-GAAP gap more than doubled year-over-year on Boyd purchase accounting and the interest cost of the $9.55B financing.

| Quarter | Adj EPS | GAAP EPS | Gap | |---|---:|---:|---:| | Q2 2026 | $3.15 | $2.11 | $1.04 | | Q1 2026 | $2.81 | $2.22 | $0.59 | | Q4 2025 | $3.33 | $2.91 | $0.42 | | Q3 2025 | $3.07 | $2.59 | $0.48 | | Q2 2025 | $2.95 | $2.51 | $0.44 |

The adjusted beat is real, but $0.60/share of incremental year-over-year add-backs is the price of it. Watch whether Q3 restores drop-through toward 1.0x — that is the single falsifiable test of the H2 story.

What was issued on 2026-07-31

| Q3 2026 (first issuance) | Low | High | Midpoint | |---|---:|---:|---:| | Organic growth | 13.5% | 15.5% | 14.5% | | Segment operating margin | 24.6% | 25.0% | 24.8% | | GAAP diluted EPS | $2.77 | $2.87 | $2.82 | | Adjusted EPS | $3.46 | $3.56 | $3.51 | | Revenue (derived) | ~$8.35B | ~$8.56B | ~$8.46B |

| FY 2026 (raised) | Low | High | Mid | Prior (Q1, 2026-05-05) | Original (Q4, 2026-02-03) | |---|---:|---:|---:|---:|---:| | Organic growth | 11.0% | 13.0% | 12.0% | 9-11% (10.0%) | 7-9% (8.0%) | | Segment operating margin | 24.1% | 24.5% | 24.3% | 24.1-24.5% (24.3%) | 24.6-25.0% (24.8%) | | GAAP diluted EPS | $10.36 | $10.56 | $10.46 | not disclosed | not disclosed | | Adjusted EPS | $13.40 | $13.60 | $13.50 | $13.05-$13.50 ($13.28) | $13.00-$13.50 ($13.25) | | Revenue (derived) | ~$32.27B | ~$32.82B | ~$32.54B | ~$31.89B | ~$30.5B |

Guidance versus consensus

| FY 2026 metric | New guide mid | Consensus | vs prior guide | vs consensus | |---|---:|---:|---:|---:| | Revenue (derived) | ~$32.54B | $32.14B | +$0.65B (+2.0%) | +1.3% | | Organic growth | 12.0% | ~11.0% implied | +200 bps | +100 bps | | Segment margin | 24.3% | not published | 0 bps | — | | Adjusted EPS | $13.50 | $13.35 | +$0.22 (+1.7%) | +$0.15 (+1.1%) |

| Q3 2026 metric | Guide mid | Consensus | vs consensus | |---|---:|---:|---:| | Revenue (derived) | ~$8.46B | $8.267B | +2.3% | | Organic growth | 14.5% | ~11.6% implied | +290 bps | | Adjusted EPS | $3.51 | $3.499 | +$0.01 — in line |

The FY26 adjusted-EPS guide bridge

The three things the headline raise hides

(a) The organic raise is real; the EPS raise is mostly the Q2 beat flowing through. Only ~$0.07-$0.12 of the $0.22 is a genuine second-half increase, against a 200 bp organic raise worth roughly $0.55B of incremental revenue. Conversion is poor because the margin guide did not move and Boyd's $9.55B of debt sits below the line.

(b) The margin guide has never been repaired and now guides down year-over-year. Trajectory: 24.6-25.0% (Feb) → cut 50 bps to 24.1-24.5% (Q1, "primarily due to Electrical Americas Q1 performance") → unchanged at Q2, despite a 200 bp organic raise and a Q2 margin that landed 10 bps above the guide high end. At the 24.3% midpoint FY2026 is 20 bps below FY2025's 24.5%; Q3 is guided to 24.8% mid versus 25.0% actual a year ago — also −20 bps.

(c) The back half is a very large step, and management pre-defended it. 1H26 segment margin was 22.91%. The FY midpoint on ~$32.54B implies 2H segment profit of $4.246B on $16.56B = 25.64%, a +272 bp H1-to-H2 step. With Q3 guided to 24.8%, Q4 is implied at ~26.5% — the highest quarterly segment margin in Eaton's history.

| Period | Organic YoY | Adj EPS | Adj EPS YoY | Segment margin | Margin YoY bps | Basis | |---|---:|---:|---:|---:|---:|---| | Q2 2025 | 8% | $2.95 | +8% | 23.9% | +20 | actual | | Q3 2025 | 7% | $3.07 | +8% | 25.0% | +70 | actual | | Q4 2025 | 9% | $3.33 | +18% | 24.9% | +20 | actual | | Q1 2026 | ~11% | $2.81 | +3.3% | 22.7% | −120 | actual | | Q2 2026 | 14% | $3.15 | +6.8% | 23.1% | −80 | actual | | Q3 2026 | 14.5% | $3.51 | +14.3% | 24.8% | −20 | guide mid | | Q4 2026 | ~9% | ~$4.03 | +21.0% | ~26.5% | +161 | derived plug |

Read. Organic revenue growth is accelerating and guidance extends the acceleration one more quarter — 7% → 9% → 11% → 14% actual, then 14.5% guided for Q3 — but the FY26 guide mid then plugs Q4 back down to ~9%, a ~550 bp step-down management was asked about on the call and did not explain. Adjusted EPS growth is the mirror image: it decelerated through 1H26 and the guide calls for a sharp re-acceleration. Margin is the hinge — reported margin is contracting year-over-year and the guide asks for that to flip to +161 bps in Q4. The revenue guide is an extrapolation of a visible trend; the EPS and margin guide is an inflection that has not yet appeared in a reported quarter. That is the asymmetry to price.

The raise itself, versus the same moment last year

| At the Q2 call | 2025 | 2026 | Change | |---|---:|---:|---| | Organic guide raise | +50 bps (to 8.5-9.5%) | +200 bps (to 11-13%) | 4x larger | | Adj EPS midpoint raise | +$0.07 (to $12.07) | +$0.22 (to $13.50) | 3x larger | | Segment margin guide raise | +10 bps (to 24.1-24.5%) | 0 bps (24.1-24.5%) | Worse | | Q2 result vs own guide | organic and EPS at the high end | organic and EPS above the high end | Better | | Next-quarter EPS guide vs consensus | — | in line (+$0.01) | Not a raise |

Track-record caveat. FY2025 organic was raised to 8.5-9.5% at the Q2 2025 call, then walked to "likely the low end" at Q3 2025, and landed at 8% — the low end. This team has raised into a second half before and had to trim. The difference now is backlog coverage (+103% Electrical Global) that did not exist in 2025.

Tone: Q1 versus Q2

| Dimension | Q1 2026 call | Q2 2026 call | Shift | |---|---|---|---| | Overall posture | Explaining a margin shortfall | Claiming momentum | Defensive → offensive | | Signature line | "strongest growth opportunities still in front of us" | strategy is working and "gathering pace" | More assertive, present tense | | Margin language | "Q1 will be the trough"; guide cut 50 bps | 190 bps sequential Americas expansion delivered; guide held, not raised | Executed, but did not re-raise | | Growth language | "demand environment remains exceptional" | "orders accelerating across all electrical end markets" | Broadened beyond data centre | | Self-restraint | "we set realistic expectations, which we aim to beat" | remains "very prudent" having raised 450 bps cumulatively | Consistent sandbagging posture | | Data centre | +50% organic (Americas) | +65% Electrical Global organic vs a market growing 23%; 8th straight quarter of 35%+ | Accelerating, now quantified vs market |

Confidence is highest on orders and backlog and lowest, by revealed behaviour, on margin. Eaton raised organic 200 bps and left the margin range untouched. That is what a management team does when it believes the volume but not yet the drop-through.

| Metric | Q3'24 | Q4'24 | Q1'25 | Q2'25 | Q3'25 | Q4'25 | Q1'26 | Q2'26 | |---|---:|---:|---:|---:|---:|---:|---:|---:| | Revenue YoY % | +7.9% | +4.6% | +7.3% | +10.7% | +10.1% | +13.1% | +16.8% | +21.4% | | Rev accel (bps QoQ) | −34 | −333 | +273 | +337 | −54 | +293 | +378 | +454 | | Adj EPS YoY % | +15.0% | +11.0% | +13.3% | +8.1% | +8.1% | +17.7% | +3.3% | +6.8% | | EPS accel (bps QoQ) | −855 | −400 | +235 | −527 | +4 | +957 | −1,436 | +347 | | EPS YoY less Rev YoY (bps) | +707 | +641 | +603 | −262 | −204 | +461 | −1,353 | −1,461 |

Why the gap exists — the bridge, quantified

| Driver | Q2'25 | Q2'26 | YoY delta | Approx. adjusted-EPS impact | |---|---:|---:|---:|---| | Interest expense | $71M | $201M | +$130M | ≈ −$0.27/sh, ≈ −900 bps of EPS growth | | Gross margin | 37.0% | 33.5% | −350 bps | Boyd mix + price/cost lag | | GAAP operating income | $1,256M | $1,392M | +10.8% | vs revenue +21.4% — negative GAAP operating leverage | | GAAP pre-tax income | $1,186M | $1,144M | −3.5% | pre-tax profit fell on record revenue | | GAAP effective tax rate | 17.1% | 28.1% | +1,100 bps | GAAP only; adjusted rate unaffected | | Diluted shares | 392.5M | 389.5M | −0.76% | ≈ +76 bps tailwind |

The interest line alone explains most of the 1,461 bps gap. Eaton borrowed to fund the $9.55B Boyd Thermal acquisition; total debt is $21.33B against $0.48B of cash, 3.44x net debt/EBITDA TTM. Interest expense has gone from $6M in Q2 2024 to $201M in Q2 2026 — a 33x increase. Management named this explicitly and in advance: on the Q1 2026 call Foster said the FY guide includes "absorbing the Boyd dilution to EPS." Known, disclosed, self-inflicted and — on management's own bridge — temporary.

Anyone underwriting ETN on GAAP earnings sees a company in decline; anyone underwriting on revenue and orders sees the opposite. That divergence is the single most important thing to be right about on this name. Our read: the EPS trough was Q1 2026 and the inflection is the quarter now in progress. The risk to that call, stated plainly, is that ~300 bps of the H2 bridge is price/cost that only fully landed in August — if it slips a quarter, FY26 lands nearer $13.00 than $13.50 and the EPS series stays flat for two more quarters while the revenue line keeps accelerating.

| # | Catalyst | Timing | What consensus believes | Q2 print / management signal | |---|---|---|---|---| | 1 | Q3 segment-margin step-up to 24.6-25.0% | 2026-11-03 | Street takes the guide at face value; Q3 consensus $3.52 vs guide mid $3.51 — zero cushion | Q2 margin 23.1%, 10 bps above the guide high end. Guide implies +150 to +190 bps QoQ. Foster: H1-to-H2 Americas expansion of 450-500 bps = ~300 bps price/cost + 150-200 bps output. All pricing implemented in Q2 or early August | | 2 | FY26 raise cadence — a fourth raise at Q3 | 2026-11-03 | FY26 consensus $13.35 sits 15c below the new $13.50 midpoint; street has not caught up | Organic guide raised +200 bps; sequence through 2026 is 8% → 10% → 12%. Ruiz still calls the company "very prudent" | | 3 | Electrical Global backlog conversion + Boyd ramp | Q3 2026 → FY2027 | FY27 consensus $35.55B (+10.6%) / $15.73 embeds a sharp organic deceleration from the 14% just printed | Backlog +103% YoY (+54% organic), book-to-bill 1.2, orders +33% R12M. Boyd $432M in Q2; FY26 Boyd raised to $1.8B ($1.5B on Eaton's books). Data-centre organic +65% vs a market growing 23% | | 4 | 800 VDC transition / NVIDIA co-design | Design wins now; revenue 2027 | Not in numbers. Street models $3.4M content/MW on today's AC architecture | Four building blocks required: MV solid-state transformers (Eaton leads post-Resilient Power), DC breakers, power electronics/UPS, cooling. Only vendor claiming all four plus one-hour service response | | 5 | Mobility separation — RMT with Dana | Close expected Q1 2027 | Modelled as a clean-up, not a re-rate; FY27 not fully rebased for deconsolidation | Mobility is the only shrinking segment: −2% organic, 13.0% margin. Organic ex-Mobility was +16% vs +14% reported. Eaton receives ~$1.1B cash; ETN holders own at least 50.1% | | 6 | Capacity ramp → lead-time compression | New plants 1H27-2027 | Street treats capacity as a cost line (~$1.15B capex), not a revenue unlock | >$1B across 24 Americas facilities (12 ramped, 6 more by year-end, 6 beyond 2027). Bellevue NE MV switchgear opens 1H27; Jonesville SC ($340M) and Richmond VA in 2027 | | 7 | Hyperscaler 2027 capex guides | Late Oct 2026 | Big-4 2026 capex ~$725B; sell-side already modelling >$1T for 2027 | Lands days before ETN's Q3 print. Eaton's orders are the derivative — Americas orders +41% R12M, backlog +33% | | 8 | Non-data-centre electrical recovery | Already inflecting | Under-modelled — the street's ETN debate is almost entirely data centre | Commercial/institutional, machine OEM and distributor IT all up double digits; orders mid-to-high teens across commercial, institutional, utility and industrial. The sentiment-inversion leg | | 9 | Aerospace | Watch Q3 margin | Low expectations; stable-margin ballast | $1.222B, +7% organic, margin 22.8% (+58 bps YoY), orders +17%, backlog +28%, book-to-bill 1.2. Organic decelerated from +9% while backlog built | | 10 | Tariffs / trade policy | Ongoing | Was a 2025-26 bear case | $2.8M — under $0.01 of EPS. Effectively retired as a topic |

Scorecard for the Q3 print (2026-11-03)

| Metric | Q2 2026 actual | Q3 2026 guide | Q3 consensus | Bar to clear | |---|---|---|---|---| | Organic growth | +14.0% | +13.5% to +15.5% | rev $8.310B (+18.9% reported) | Guide mid 14.5% = +50 bps acceleration vs Q2 | | Segment margin | 23.1% | 24.6% - 25.0% | not published | +150 to +190 bps QoQ — the whole H2 EPS bridge | | Adjusted EPS | $3.15 | $3.46 - $3.56 | $3.52 | Consensus is 1c above the midpoint — no slack | | FY26 adj EPS | — | $13.40 - $13.60 | $13.35 | Consensus 15c below the guide midpoint |

The Q3 setup is unusually binary. Consensus sits above the guide midpoint, so ETN has to beat its own guide again — which it has done in 9 of the last 9 quarters, with magnitude widening from +$0.01/+$0.02 through 2025 to +$0.08 in each of Q1 and Q2 2026.

Calendar

| Date | Event | Why it matters | |---|---|---| | Early Aug 2026 | Remaining pricing actions land | Completes the ~300 bps price/cost leg of the H2 margin bridge | | Late Oct 2026 | Hyperscaler Q3 prints, first hard 2027 capex frames | Sets the FY27 orders narrative days before ETN reports | | 2026-11-03 | ETN Q3 2026 earnings | Margin step-up proof point; fourth FY26 raise if the pattern holds | | Q4 2026 | Dana S-4 / shareholder vote | Gating item for the RMT | | Q1 2027 | Mobility / Dana RMT close | ~$1.1B cash in; removes the −2% organic drag | | 1H 2027 | Bellevue NE MV switchgear plant opens | Lead-time relief in the most constrained product line | | 2027 | Richmond VA + Jonesville SC production; 800 VDC deployments | Content-per-MW step-up above the $3.4M assumption |

What would break the catalyst set

| Risk | Trigger to watch | |---|---| | Margin step-up fails | Q3 segment margin under 24.6%; price/cost realisation slipping past August | | Backlog quality | Book-to-bill falling below 1.0 in Electrical Global; orders decelerating from +33% / +41% | | Hyperscaler digestion | Any big-4 2027 capex guide framed flat-to-down in late-October prints | | Boyd | Boyd tracking below the raised $1.8B; GAAP-to-adjusted gap widening further | | Leverage | Net debt/EBITDA 3.44x TTM post-Boyd — limits M&A optionality until the $1.1B Dana proceeds land | | RMT slippage | Dana vote or regulatory delay pushing the close past Q1 2027 |

A strong Q&A performance on the quarter and a weak one on the forward setup. Every question about what already happened — segment growth, the Q1-to-Q2 margin bridge, Boyd's contribution, tariffs — got a quantified answer, several unusually precise. All three deflections cluster on the same theme: when the demand converts and what the constraint is.

| # | Analyst | Firm | Topic | Badge | |---|---|---|---|---| | 1 | Deane Dray | RBC | Non-data-centre electrical demand | Well answered | | 2 | Andrew Obin | Bank of America | Boyd advantages and cadence | Well answered | | 3 | Chris Snyder | Morgan Stanley | Electrical Global upside and trajectory | Well answered | | 4 | Scott Davis | Melius | CEO's first year | Well answered (low-information question) | | 5 | Nigel Coe | Wolfe | Tariffs and the Americas margin bridge | Well answered — best on the call | | 6 | Jeffrey Sprague | Vertical | Does volume support the margin ramp? | Well answered (assertion-led) | | 7 | Nicole DeBlase | Deutsche Bank | Implied back-half deceleration | Deflected | | 8 | Andrew Kaplowitz | Citi | Content per megawatt and 800V DC | Well answered | | 9 | Chad Dillard | Bernstein | Prefab/modular and 800V adoption | Deflected | | 10 | Jeff Hammond | KeyBanc | Lead times as capacity comes online | Deflected |

The best answer on the call — Coe on the Americas margin bridge

Tariff impact $2.8M, under $0.01 of EPS. The Q1-to-Q2 sequential +190 bps breaks into ~100 bps price/cost and ~90 bps output scaling. The H1-to-H2 guide implies 450-500 bps of further expansion: ~300 bps price/cost, 150-200 bps output/productivity, with all pricing actions implemented in Q2 or early August. This exact question was asked on the Q1 2026 call and answered qualitatively; management has now delivered the promised sequential recovery and upgraded the disclosure from narrative to a quantified bridge. Promise made, promise kept, disclosure improved — the single most management-quality-positive item in the quarter.

The three deflections

| # | Analyst | What was asked | What came back | Why it matters | |---|---|---|---|---| | 7 | Nicole DeBlase | Why does the guide embed a Q4 step-down | "Very prudent"; the 8% → 10% → 12% raise trajectory | FY midpoint implies Q4 organic ~+9% vs the Q3 guide's 14.5% mid — a ~530 bps deceleration, unexplained. Sensitivity: the 11% low end implies Q4 ~+5%; the 13% high end ~+13% | | 9 | Chad Dillard | 800V adoption progress; prefab strategy | Architecture and partnership language; no milestone update | On the Q1 call Ruiz committed to "orders in the second half of the year" for 800V. Q2 is the first H2 print and produced no order count, no dollar value, no restated timeline. A missed self-set milestone that goes unmentioned is worse than one missed and explained | | 10 | Jeff Hammond | How much lead times improve as capacity lands | Eight quarters of 35%+ data-centre growth; "improvement trajectory" | A demand statistic answering a supply question. No lead time in weeks, for any product, at any point in time. With backlog +103% and book-to-bill 1.2, lead time is the conversion variable |

The pattern is coherent, not three separate lapses. Every deflection substitutes an order/backlog/growth number for the capacity, timing or lead-time number actually requested. That is precisely the posture a company adopts when supply, not demand, is the binding constraint and it does not yet want to be held to a supply schedule. Calibration: this does not read as evasion of bad news — management raised twice this year, beat by $0.08, delivered the Americas margin recovery they promised, and volunteered the ex-Boyd backlog deflator unprompted. The withheld items are things they are not yet confident enough to be scored on.

Blind spots — what the street did not ask at all

| Topic | The figure | Why it should have been asked | |---|---|---| | GAAP vs adjusted divergence | GAAP diluted EPS $2.11 vs $2.51 (−15.9% YoY) while adjusted rose to $3.15. The wedge widened to $1.04 from $0.44 | GAAP earnings fell in a record quarter, and FY26 GAAP guidance is flat at the midpoint against adjusted guided +12%. Nobody asked what sits in the $1.04 | | Gross margin compression | 33.5% vs 37.0% — the fifth consecutive quarter of year-over-year compression | Segment margin beat guidance so the headline looked fine, but gross margin has fallen ~540 bps from the peak. The H2 margin thesis lives or dies here | | Leverage post-Boyd | Net debt / EBITDA 3.44x (debt $21.33B, cash $0.48B) after the $9.55B acquisition | No deleveraging path, target, or capital-allocation priority discussed | | Mobility separation | RMT with Dana closing Q1 2027; Mobility ran −2% organic at 13.0% margin | A structural transaction removing a full segment drew no questions on terms, proceeds, stranded cost or post-separation margin |

Carry-forward for the Q3 call

1. Have the 800-volt DC orders promised for H2 2026 been booked? Count and value.
2. Quote a lead time — weeks, by major product line, versus a year ago.
3. Explain the Q4 organic step-down to ~+9% implied, or raise the FY guide again and retire the question.
4. Size the $1.04 GAAP-to-adjusted wedge and the path back for GAAP EPS.
5. Boyd standalone margin and accretion — revenue has been disclosed three times; profitability, never.

Eaton's management is generally accurate and consistent on the things it can control. The contradictions cluster in exactly two places, and both are where the bull case lives: the Boyd / data-centre TAM math, and the Electrical Americas capacity-ramp and price/cost story. Every broken claim below is a forward or definitional claim; every claim that held is an arithmetic one about delivered results.

Bottom line on the contradictions. The two that should change how the quarter is underwritten are C1 (the Boyd beat is overstated by 12 points against Eaton's own published Q2 assumption) and C2 (the content-per-megawatt multiplier every sell-side data-centre TAM runs on was restated up 21% pre-Boyd with no bridge, and reaffirmed this quarter at the higher level). The two that should change how the guidance is underwritten are C4 (half the incremental capacity slipped out of the 2H 2026 window the margin bridge depends on) and C5 (price/cost was declared solved two quarters before it became two-thirds of the H2 recovery).

None of this impeaches the demand signal, which is corroborated by numbers management cannot restate: backlog +103%, orders +41%, book-to-bill 1.2, and nine consecutive adjusted-EPS beats with the magnitude widening. The contradictions are concentrated on the margin and TAM narrative layered on top of that demand — which is precisely the layer where the multiple is being paid (~31.1x FY26E, ~26.4x FY27E consensus EPS).

The three macro datapoints that matter

1. US data-centre backlog went from 228 GW to 307 GW in one quarter. Management framed it as 15 years of backlog at 2025 build rates, up from 12 years one quarter earlier. That is a +79 GW single-quarter increase in announced backlog while the market debates an AI capex digestion phase. Either the hyperscaler capex cycle has a much longer tail than consensus models carry, or the backlog is being padded with speculative announcements. Eaton is voting hard for the former with more than $1B of capex. Corroborating: Electrical Global data-centre organic revenue +65% against a market Eaton sizes at +23% — roughly 42 points of claimed share gain — and an eighth consecutive quarter of 35%+ growth.

2. Machine OEM orders inflected to the mid-30s%, and nobody followed up. Machine OEM is the classic early-cycle electrical read — component demand sitting inside somebody else's capital equipment. A mid-30s order rate off a base that was declining through 2025 says the short-cycle industrial destock is over and the restock has begun; the distributor/IT channel returning to double-digit growth is the confirming signal. Note the spread: orders mid-30s versus revenue double digits, roughly 2x — a leading-indicator gap, not a delivered result. That is the definition of a read-through the street is not pricing, and it went unasked because every analyst wanted to talk about data centres.

3. The H2 margin bridge is 60% price, not volume — ~300 of 450-500 bps, with all pricing actions already implemented. That converts the H2 guide from a demand bet into an execution bet, which is a higher-confidence setup. But it also means the whole group's H2 margin story hinges on price stickiness rather than growth. Anyone in the electrical complex whose H2 bridge depends on volume is running a lower-quality bridge than Eaton's.

Other macro reads

| Topic | What was said | Read-through | |---|---|---| | Interest rates | Nothing. Zero mentions of rates, the Fed or cost of capital in prepared remarks or Q&A | The absence is the signal: Eaton's demand is backlog-governed, not rate-governed. Industrial names still framing 2026 around rate cuts do not share that insulation | | Tariffs | $2.8M in Q2, under $0.01 of EPS; H2 impact "immaterial" and already embedded | A US-heavy, local-for-local footprint is a demonstrated tariff hedge. Removes tariffs as an excuse line for anyone in the group who misses in H2 | | Labour scarcity | "There is scarcity of people to work on a stick build" — electricians, plumbers | The binding limit on US data-centre build is increasingly skilled trade labour, not equipment or capital. Durable share shift toward prefabricated/modular content | | Residential | "not as strong as you guys know" | The sole soft end market, and a step down from Q1 where residential was cited as a Global strength. Small for Eaton, more meaningful for resi-levered electricals | | Consumer | Mobility organic −2%, mostly a deliberate exit of low-margin North America light vehicle | Consumer/housing channel is stable-weak, not deteriorating. Consistent with a bifurcated economy — infrastructure/AI capex booming, household-facing demand flat |

Companies and cohorts

| Company / cohort | Type | Datapoint | Better / worse for whom | |---|---|---|---| | Dana Incorporated (DAN) | Named | RMT; ETN holders own at least 50.1%; Eaton receives ~$1.1B cash before close; Mobility valued ~$5.1B | Better for ETN, mixed-to-worse for DAN. Eaton exits ~$3B of revenue at 13.0% margin and takes $1.1B out first, leaving a business growing 25%+ in Electrical at 20-27% margins. Watch DAN's pro-forma leverage disclosure | | Boyd Thermal (ETN-owned) | Named | $432M in Q2; FY26 raised to $1.8B ($1.5B on Eaton's books) | Better for the liquid-cooling complex. A $432M quarter one quarter after close is a demand statement, not an integration statement. Anyone modelling liquid cooling as a 2027-28 ramp is late | | Sub-scale cooling / power specialists | Implied | "some are actually showing some quality issues from the get-go" — Ruiz | Worse. The most aggressive competitive line on the call, and a quality accusation rather than a price one. Share consolidating toward whoever can scale at hyperscaler quality standards. Checkable against peer warranty/rework disclosures next quarter | | NVIDIA (NVDA) | Implied in Q2, named in Q1 | Boyd is "a design partner for chip providers" with "first look and first bid on future programs"; Q1 named the Vera Rubin partnership as "already 800-volt DC" | Better for the rack roadmap. Note the downgrade in specificity — NVIDIA named repeatedly in Q1, only "chip providers" in Q2. Read as broadening rather than cooling, but monitor | | Hyperscalers and colos (MSFT, AMZN, GOOGL, META, ORCL; DLR, EQIX) | Implied | 307 GW / 15 years of US backlog, up from 228 GW / 12 years | Better for the capex trade; worse for the AI-digestion thesis | | Resilient Power (ETN-owned) | Named | "we are clearly ahead in terms of technology after the acquisition of Resilient Power" on MV solid-state transformers | Worse for legacy AC transformer / switchgear incumbents (Hitachi Energy, Schneider, ABB, Siemens Energy) if the DC transition sticks. Also erodes the scarcity value of merchant medium-voltage transformers | | Fibrebond (ETN-owned) | Named | Named in direct answer to the electrician-scarcity question | Better for modular/prefab; incrementally worse for stick-build contractors (PWR, EME, IES, FIX) at the margin. Does not shrink near-term backlogs, but caps the pricing power they extract from labour scarcity | | Ultra PCS (ETN-owned) | Named | "performing to our expectations" — Foster | Better for defence electronics and aftermarket, but "in line" is a downgrade in tone versus Boyd's "20% above." Aerospace organic decelerated +9% to +7% while orders +17% and backlog +28% — demand fine, conversion constrained | | Machine OEM cohort (CAT, DE, Terex, AGCO, Oshkosh) | Implied | Orders "in the mid-30s"; revenue double digits | Better for short-cycle industrial machinery — the cleanest early-cycle read on the call | | Electrical distributors (WCC, GWW, Rexel, Sonepar) | Implied | "distributor IT recovering really nicely, also double digits" | Better for WESCO and the distribution channel — inventory cleared, and both volume and price tailwinds in H2 | | Utility / T&D capex complex | Implied | "strength in data center utility and machine OEM"; utility inside the mid-to-high-teens order cohort | Better for the utility T&D capex cycle. Utility now growing alongside data centres rather than as a slower separate bucket | | Residential chain (homebuilders, Generac, Hubbell resi) | Implied | "even residential, which is a market that is not as strong as you guys know" | Worse — the only explicitly weak end market Eaton named | | Aerospace OEMs (Boeing, Airbus) | Implied | +7% organic, orders +17%, backlog +28% | Neutral-to-slightly-worse. Eaton shipping below order rates — either OEM build rates are gated or Eaton is capacity-constrained in aerospace too. Cross-check against delivery cadence | | Siemens Energy | Named on the Q1 call, not repeated in Q2 | Q1: "the on-site power partnership with Siemens Energy to help solve for global power constraints" | Better for on-site / behind-the-meter generation. Not re-mentioned in Q2, so treat as unchanged. Grid interconnect queues pushing data centres to bring their own generation remains the key call option for turbine and gen-set suppliers |

1. Q3 segment margin against the 24.6-25.0% guide — the single cleanest test of the entire H2 bridge.
2. Electrical Americas margin in the 29-30% range to keep the "exit the year above 30%" commitment live.
3. Whether the FY26 segment-margin range finally moves up off 24.1-24.5%, where it has sat since the Q1 cut.
4. Drop-through ratio recovering toward 1.0x from 0.57x.
5. 800-volt DC orders — count and value, against the H2 2026 commitment made on the Q1 call.
6. A lead time quoted in weeks, by major product line, versus a year ago.
7. The implied ~+9% Q4 organic step-down explained, or a fourth raise that retires the question.
8. The $1.04 GAAP-to-adjusted wedge sized, with a path back for GAAP EPS guided flat for FY26.
9. Boyd standalone margin and accretion — revenue disclosed three times, profitability never.
10. Any update to the FY26 free-cash-flow guide of $3.9-4.3B, not restated in the Q2 materials.