A boring enclosures-and-connectors company that bought and sold its way into the AI data-center build — now growing 50%+ with infrastructure orders up 100% organically; the re-rating to ~28x is largely earned, but the multiple now prices in the build continuing, and the legacy two-thirds of the business is still cyclical late-cycle industrials.
| Date |
|---|
| Type |
|---|
| What happened |
|---|
| Source |
|---|
| 2026-08-10 | editorial note | Capex figure revised: 40% → $130MCapex moved from 40% (deep-dive-2026-06-20.md) to $130M (deep-dive-2026-08-10.md). | dossier |
| 2026-08-10 | editorial note | Margin figure revised: $80M → 24.4%Margin moved from $80M (deep-dive-2026-06-20.md) to 24.4% (deep-dive-2026-08-10.md). | dossier |
| 2026-08-10 | editorial note | Revenue figure revised: $3.9B → 19%Revenue moved from $3.9B (deep-dive-2026-06-20.md) to 19% (deep-dive-2026-08-10.md). | dossier |
| 2026-08-10 | editorial note | Verdict changed: A boring enclosures-and-connectors company that bought and sold its way into the AI data-center build — now growing 50%+ with infrastructure orders up 100% organically; the re-rating…Before (deep-dive-2026-06-20.md): A boring enclosures-and-connectors company that bought and sold its way into the AI data-center build — now growing 50%+ with infrastructure orders up 100% organically; the re-rating to ~28x is largely earned, but the multiple now prices in the build continuing, and the legacy two-thirds of the business is still cyclical late-cycle industrials. After (deep-dive-2026-08-10.md): The order line decelerated and the backlog shrank — and guidance still jumped 12% while the stock went nowhere. nVent shipped its backlog instead of hoarding it, forward earnings power rose ~17%, and the multiple came down to meet it; the entry is materially better at $165 in August than it was at $167 in June, but the "booked, not hoped" leg of the bull case is now unproven rather than proven. | dossier |
The verdict
The order line decelerated and the backlog shrank — and guidance still jumped 12% while the stock went nowhere. nVent shipped its backlog instead of hoarding it, forward earnings power rose ~17%, and the multiple came down to meet it; the entry is materially better at $165 in August than it was at $167 in June, but the "booked, not hoped" leg of the bull case is now unproven rather than proven.
Primary sources
SEC filings
Earnings calls
Source documents — open to read in full
Q2-2026 was a blowout, and bigger than Q1's blowout. Net sales $1,471.3M, +52.8% (organic +46.9%); adjusted EPS $1.45, +69%; GAAP diluted continuing EPS $1.32, +103%. The adjusted number beat consensus by ~25% and revenue by ~19% — a wider beat than Q1's ~16%. ``
Guidance raised a third time, by 12%. FY2026 adjusted EPS to $5.00–$5.10 from $4.45–$4.55; organic sales growth to 32–34% from 21–23%; reported to 37–39% from 26–28%. At the $5.05 midpoint that is +50% on FY2025's $3.35. ``
Orders decelerated hard and the backlog fell. Organic orders +low-double-digits — down from +~40% in Q1 — and backlog $2.5B, down from $2.6B at Q1. This is the first sequential backlog decline of the AI cycle. Management framed it as lumpiness plus deliberate backlog conversion ("we worked hard in Q2 to really execute on that backlog") and volunteered that Q3 orders have been "very strong thus far." ``
First hard data-center number, ever: $2B of FY2026 sales, >2x FY2025. nVent has never quantified the vertical before. Against guided FY2026 sales of ~$5.37B that is ~37% of the company . Infrastructure as a whole went **12% of sales at spin → 45% in FY2025 → ~60% in H1-2026**.
A third liquid-cooling plant, and a 40% capex step-up. "Blaine 2" — 160,000 sq ft in Minnesota, production H1-2027, 200+ jobs — is the third liquid-cooling expansion in three years (>400,000 sq ft added total). FY2026 Capital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. guided to ~$130M, +40% YoY. ``
The stock round-tripped and ended flat while earnings power rose. $167.34 (6/16) → a new ATH of $184.64 intra-window → $141.75 on 2026-07-28 (−5.0% that day, −12.9% on the month, in a sector-wide AI-infra valuation reset) → +7.78% to $156.07 on the Q2 print → $164.89 on 2026-08-10. Net −1.5% across the refresh window against a +12% FY2026 EPS raise. ``
The structural thesis stands, and one leg of it strengthened. nVent is still the mid-cap electrical hardware supplier levered to the AI data hall through both gray space and white space; the mix keeps tilting toward the leveraged half (Systems Protection now 72.9% of Q2 sales vs 72.0% in Q1 and 66.6% in FY2025 ``); management is still executing.
Two prior-dossier open questions got answered, both favourably:
What did not hold: the quality of the growth evidence. The prior dossier's single strongest bull sentence was that "orders are outrunning revenue, so the growth is booked, not hoped." That is no longer true this quarter — revenue outran orders and the backlog shrank. The growth is now shipped, which is better cash-wise and worse visibility-wise. This is the material degradation of the refresh and it is why Lens 12/13 change shape below.
Lens 1 (Company overview) — carried. See the previous dossier. One correction worth stamping: the CFO is Gary L. Corona, EVP & CFO (signatory on both 8-Ks in the window) `` — the prior dossier named the CEO but not the CFO.
Lens 2 (Supply chain) — unchanged in structure, one number moved. See the previous dossier. The tariff drag is now ~$100M for FY2026, up from ~$80M; management attributes the increase primarily to higher volume, not a worse tariff rate, and reiterates full offset via price, supply-chain productivity and mitigation ``. Q2 alone absorbed >$50M of inflation including >$30M of tariff impact, fully offset by price plus productivity — the proof is the +110bps adjusted ROS expansion. A genuinely new wrinkle: $25.8M of IEEPA tariffs previously remitted were reimbursed in Q2 (see Lens 10).
Lens 3 (Moats) — carried. See the previous dossier. One data point cuts toward the moat being wider than the prior pass allowed: new products contributed >30 points of sales growth in Q2 with 14 launched in the quarter, and a modular liquid-cooling platform launches this autumn designed hot-swappable with an attached service capability. If that lands, the "commoditizable metal-basher" bear leg weakens — modularity plus service is exactly the content-per-rack defence the prior dossier said was missing. It is a claim, not yet a result. ``
All and.
Net sales ($M):
| Segment | Q2-2025 | Q2-2026 | Δ | organic | H1-2026 | Δ |
|---|---|---|---|---|---|---|
| Systems Protection | 632.0 | 1,072.1 | +69.6% | +62.0% | 1,966.9 | +72.5% |
| Electrical Connections | 331.1 | 399.2 | +20.6% | +17.9% | 746.4 | +18.1% |
| Total | 963.1 | 1,471.3 | +52.8% | +46.9% | 2,713.3 | +53.1% |
Reportable segment income ($M) and return on sales:
| Segment | Q2-2025 | Q2-2026 | Δ | ROS Q2-25 | ROS Q2-26 | Δ bps |
|---|---|---|---|---|---|---|
| Systems Protection | 137.1 | 248.2 | +81.0% | 21.7% | 23.2% | +150 |
| Electrical Connections | 94.9 | 108.8 | +14.6% | 28.7% | 27.3% | −140 |
| Reportable total | 232.0 | 357.0 | +53.9% | — | — | — |
| Enterprise & other | (32.0) | (34.3) | — | — | — | — |
| Adjusted operating income | 200.0 | 322.7 | +61.4% | 20.8% | 21.9% | +110 |
Three reads.
Vertical / geography (Q2, organic): infrastructure more than doubled; power utilities double-digit; commercial & residential high-single-digit; industrial low-single-digit. Americas very strong double-digit, Europe mid-single, Asia-Pacific double-digit. ``
The number. Net sales $1,471.3M, +52.8% YoY (organic +46.9%, acquisition +5.4pts, FX +0.5pts). Gross profit $558.0M (37.9%, −70bps). Operating income $300.7M (20.4%, +410bps). Net income from continuing operations $215.8M. GAAP diluted continuing EPS $1.32 (+103%). ``
Versus consensus. Adjusted EPS $1.45 beat by ~25%; revenue beat by ~19%. `` This is the third consecutive large beat and the largest of the three.
The honest margin line. The +410bps GAAP ROS improvement is partly a one-off: it includes $25.8M of IEEPA tariff reimbursements which management explicitly excludes from adjusted results. The clean number is adjusted operating income $322.7M, adjusted ROS 21.9%, +110bps — real, volume-driven, and achieved while absorbing >$50M of inflation. Anyone quoting +410bps is quoting a flattered figure. ``
Guidance — raised for the third time. ``
| Metric | Prior guide | New guide (2026-07-31) | Δ |
|---|---|---|---|
| FY2026 adjusted EPS | $4.45–$4.55 | $5.00–$5.10 | +12.2% at midpoint |
| FY2026 GAAP EPS | $3.68–$3.78 | $4.29–$4.39 | +16.4% |
| FY2026 reported sales growth | 26–28% | 37–39% | +11pts |
| FY2026 organic sales growth | 21–23% | 32–34% | +11pts |
| Q3-2026 adjusted EPS | — | $1.35–$1.38 | +50% YoY at midpoint |
| Q3-2026 organic sales growth | — | 32–35% | reported = organic |
| FY2026 tariff impact | ~$80M | ~$100M | worse, on volume |
| FY2026 capex | — | ~$130M | +40% YoY |
The number nobody put on a slide. H1 delivered $2.54 of adjusted EPS ($1.09 + $1.45); Q3 is guided to $1.365 at midpoint; the FY midpoint is $5.05. That implies Q4-2026 adjusted EPS of ~$1.145 — **below Q3 and below Q2**. Jeffrey Sprague (Vertical Research) put this to management directly; the answer was that it is "prudent" planning around the ramp of two new facilities, not a demand signal. Take it as either sandbagging (the pattern all year) or a genuine capacity-staging constraint. It is the single most falsifiable line in the guide.
Orders and backlog — the deterioration. Organic orders +low-double-digits (Q1: +~40%). Backlog $2.5B, down from $2.6B, "giving us visibility through the year and into 2027." Management's defence is that data-center orders are large and lumpy, that Q3 orders to date have been "very strong," and that the backlog fell partly because they converted it — "we worked hard in Q2 to really execute on that backlog because we know it is important to have good lead times." Asked whether $2.5B is the right level, Wozniak: "it is around the right level." ``
Both readings are internally consistent with the print. Shipping $1.47B out of a $2.6B backlog and replacing most of it is what a flat-to-slightly-down backlog looks like in a quarter with 47% organic revenue growth. But "orders +40%" was the load-bearing evidence in the prior dossier, and it is gone for now.
Balance-sheet flags ``:
| Item | 2026-06-30 | 2025-12-31 | Read |
|---|---|---|---|
| Cash | $256.0M | $237.5M | — |
| Total debt | $1,492.4M | $1,559.8M | $68.3M term-loan prepaid |
| Net debt | $1,236.4M | $1,322.3M | net leverage 1.2x vs 2.0–2.5x target |
| Receivables | $969.3M | $693.0M | DSO 60d vs 69d a year ago — improved |
| Inventories | $522.4M | $471.9M | DIO 52d vs 64d — improved |
| Payables | $496.0M | $358.9M | DPO ~49d vs 48d |
| Total equity | $3,986.9M | $3,730.2M | — |
| Goodwill + intangibles | $4,469.6M | $4,554.5M | 62.5% of total assets; tangible book ≈ −$483M `` |
Free cash flow $167.3M in Q2 (+125%), $221.1M in H1 (+86.6%); FY conversion guided at 90–95%. Covenant compliance confirmed; $600M revolver undrawn. ``
Market reaction. +7.78% to $156.07 on the print (2026-07-31), off a base already down 12.9% on the month. $164.89 on 2026-08-10; Market capitalisationThe share price multiplied by the number of shares. What the market says the equity is worth. $26.69B; 161.86M shares; trailing P/E 45.22, forward 28.57; 52-week range $85.72–$184.64. ``
Unusual vs its own history. Four consecutive quarters above $1B in sales; 47% organic in a company whose historical band was 5–10%; a segment that just added 150bps of margin while growing 70%. This is still a regime, not a quarter — and still one that can reverse.
The prior dossier's Lens 6 was web-only. transcripts/2026-q2.md is now on disk, so the Q2 read is primary; the Q3-2025 → Q1-2026 arc remains `` from the prior pass.
The tone shift, in one sentence: management stopped selling the order book and started selling execution and capacity.
| Q3-2025 → Q4-2025 | Q1-2026 | Q2-2026 | |
|---|---|---|---|
| Headline proof point | "data centers a growth vertical" | "record orders, +40% organic orders, visibility into 2027" | "record sales and EPS, well ahead of guidance"; orders +low-double-digits |
| What they lead with | demand | bookings | shipments + capacity |
| New disclosure | — | backlog $2.6B | $2B data-center sales target; infrastructure ~60% of H1 sales |
| Hedge language | durability of DC demand | (dropped) | "orders tend to be large and lumpy"; Q4 planning "prudent" |
Four things genuinely new on this call.
What to listen for next. (a) Does the organic order rate return to at least mid-teens in Q3, as management implied? (b) Does Q4 land at the implied ~$1.145, above it (sandbagging confirmed), or below (ramp constraint real)? (c) Uptake on the autumn modular liquid-cooling launch — the first direct test of whether nVent can hold premium white-space content. (d) Any restatement of customer concentration now that neo-clouds and colos are in the mix.
The July 2026 drawdown was sector-wide, not NVT-specific: on 2026-07-28 Vertiv −6.3%, Eaton −3.1%, nVent −4.7% on the same day; Vertiv had fallen ~25% in a month. The characterisation from the coverage is that this was "a repricing of valuation, financing costs and AI-return risk," not evidence of demand collapse. ``
| Company | Ticker | Mkt cap | Fwd P/E | EV/EBITDA | Note |
|---|---|---|---|---|---|
| nVent | NVT | $26.69B `` | 28.6x · **32.7x** on FY2026 guide | ~26.3x LTM adj `` | trailing P/E 45.2x |
| Vertiv | VRT | n/a (2026-08) | ~46–49x `` | n/a | trailing ~75–80x `` |
| Eaton | ETN | n/a (2026-08) | ~33x `` | n/a | trailing >38x `` |
| Schneider | SU.PA | n/a | ~30x `` | n/a | — |
| Hubbell | HUBB | n/a (2026-08) | n/a (2026-08) | n/a | prior pass: 24.9x / 19.6x at 2026-06 |
``
``
Read, and it is a real change from June. In June NVT traded at ~28–34x forward against Vertiv 44x / Eaton 32x / Hubbell 25x. Today it trades at 28.6x forward against a Vertiv that has fallen to ~46x and an Eaton at ~33x — i.e. NVT is now cheaper than Eaton on forward earnings while growing organic revenue at 32–34% versus Eaton's high-single-digits. The prior dossier's central valuation fact — "priced as a credible data-center play but not as a pure-play" — still holds, but the gap the bull case needs closed is now being closed by earnings, not by multiple. Comps coverage is thinner than I would like: I could not source current-dated market caps, EV/EBITDA or dividend yields for VRT/ETN/HUBB, and the Vertiv/Eaton multiples above are 1–4 months stale. Treat this table as directional, not as a relative-value verdict.
`` throughout.
| Date | Move | What |
|---|---|---|
| late Jun / early Jul 2026 | new ATH $184.64 | continuation of the Q1 beat-and-raise re-rating; exact date not sourced (from the 52-wk high on stockanalysis.com, 2026-08-10) |
| 2026-07-28 | −5.0% to $141.75 | sector-wide AI-infra valuation reset. 1-week −11.7%, 1-month −12.9%. Vertiv −6.3%, Eaton −3.1% the same day `` |
| 2026-07-31 | +7.78% to $156.07 | Q2 beat (+25% on adjusted EPS) and the 12% guidance raise `` |
| 2026-07-31 | (same day) | Blaine 2 announced — 160,000 sq ft, third liquid-cooling expansion in 3 years, production H1-2027, 200+ jobs `` |
| 2026-08-05 | — | CEO Wozniak exercised 46,261 options at $25.34 and sold at weighted-average $162.71 / $163.61 / $164.73 / $165.34 `` |
| 2026-08-10 | $164.89 | +0.12% on the day; market cap $26.69B `` |
Analyst positioning: 17 analysts, consensus "Strong Buy," average 12-month target $202.13 (+22.7% from spot). Goldman Sachs to $200 (from $191), Buy. Seaport Research to $215 (from $190), Buy. ``
What this window taught about the market's reaction function — and it is a genuinely new lesson. The prior dossier concluded the market reacts to "the data-center order book and guidance revisions." Half of that is now falsified: the order book decelerated sharply and the stock went up 7.78% anyway. What actually moved the tape was the guidance revision and the shipped revenue. NVT is being priced off delivered AI-infrastructure earnings, not off bookings — and, critically, it fell 12.9% into the print on a sector multiple reset that had nothing to do with nVent's own numbers. That is the signature of a name that trades as a beta expression of the AI-capex theme, which cuts exactly as hard on the way down.
Officer change (new in the window). On 2026-06-11 Randolph A. Wacker, SVP, Chief Accounting Officer and Treasurer, notified the company he intends to retire effective 2026-09-01. On 2026-06-12 the board appointed Tyler Krutzig (age 39) SVP & Chief Accounting Officer effective the same date — Assistant Corporate Controller since 2019, previously corporate controlling and external reporting at Pentair (2016–18) and audit at Deloitte & Touche (2008–16); receives the standard executive KEESA. ``
Read: routine internal succession, not a red flag — a long-tenured internal promote with the right background, announced with a full quarter of RunwayHow long the cash lasts at the current rate of spending. It shortens the moment spending rises, which is why a figure taken from a quiet quarter flatters. and a normal severance agreement. Worth noting only because the accounting chief turns over during a period of heavy percentage-of-completion project revenue and unfinalised acquisition accounting (the EPG purchase-price allocation was still "preliminary… subject to further refinement" as of Q1-2026 ``). Monitor, do not alarm.
Capital allocation — the one thing that changed and deserves a hard look.
| Use | H1-2025 | H1-2026 | Read |
|---|---|---|---|
| Share repurchases | $253.1M (4.8M sh) | $50.4M (0.4M sh) | −80% |
| Dividends | $65.7M | $68.2M | +5% per-share |
| Capex | $38.0M | $57.6M | +52%; FY guided ~$130M, +40% |
| Debt repayment | $866.3M | $68.3M | deleveraging done; net leverage now 1.2x vs 2.0–2.5x target |
``
They did not buy the dip. The Q2 repurchase table shows 7,635 shares in the April 1–25 window at $130.45, with column (c) — shares purchased under the publicly announced plan — at zero, and the $96,520,092 remaining under the 2024 authorization unchanged from 2026-03-31 through 2026-06-30. In other words: essentially no open-market buyback in a quarter when the stock traded from ~$180 down to ~$142, the cheapest it has been all year, from a balance sheet running at 1.2x against a 2.0–2.5x target. ``
Two defensible readings. Charitable: capital went into capacity (Blaine 2, +40% capex) at a moment when a dollar of plant returns far more than a dollar of stock, and the board did stack a fresh $500M authorization on 2026-05-16 (effective 2026-07-23, expiring 2029). Uncharitable: the company that repurchased $253M at ~$53/share average in H1-2025 declined to repurchase at $142 in H1-2026, which is a worse price-discipline record than it looks, and the new authorization is a gesture that costs nothing until it is used. The honest verdict: capacity over buyback is the right call on the merits; the timing is still a miss, and the next four quarters of repurchase behaviour are the test.
Insider alignment — unchanged and still the soft flag. CEO Wozniak exercised 46,261 options at $25.34 and sold into the post-print strength on 2026-08-05 at $162.71–$165.34 . This is consistent with the prior dossier's finding of low insider ownership (~0.94%) and net selling. Option-exercise-and-sell four business days after a record print is legal, common, and completely unhelpful for the "skin in the game" question. **No updated insider-ownership percentage was sourced in this window — the ~0.94% figure remains the June number and is now stale.**
Everything else carries. Wozniak's track record, archetype (professional operator / portfolio-shaper) and the Thermal-sale-into-EPG capital-allocation call are unchanged — and the Q2 print makes that call look better, not worse: EPG "continued to exceed expectations, growing sales strong double digits year over year" ``.
`` unless noted.
Resolved since the prior dossier:
New in this window:
Improving:
Regulatory findings (required).
regulatory/regulatory-findings.md was regenerated 2026-08-10 covering 2021-08-10 → 2026-08-10 via SEC EDGAR EFTS: 0 Litigation Releases, 0 AAERs naming nVent Electric. Unchanged from the prior pass. ``Model integrity — read this before any number below. model.xlsx was rebuilt on 2026-08-10 from the newly populated CSVs. Two honest flags:
our model returns "No computed values" for every output — value per share, Enterprise valueWhat it would cost to buy the whole company: its market value plus its debt, minus the cash you would get with it. Often very different from market cap., FY+3 EPS, the balance check, WACC, share price. It needs to be opened in Excel and saved before any of it computes.Therefore: no model output is cited anywhere in this dossier. Everything below is `` with the arithmetic shown, anchored on management guidance and the reported statements.
FY2026 base = $5.05 — the guide midpoint, and unusually high-confidence: $2.54 of it is already in the bank (Q1 $1.09 + Q2 $1.45), Q3 is guided to $1.365, leaving ~$1.145 to be earned in Q4. ``
FY2027 base = $6.20. ``
FY2028 base = $7.45. ``
Path summary (adjusted EPS, $):
| Case | FY2026 | FY2027 | FY2028 |
|---|---|---|---|
| Bull | 5.15 | 7.15 | 9.10 |
| Base | 5.05 | 6.20 | 7.45 |
| Bear | 4.95 | 5.08 | 4.43 |
Versus the prior dossier (2026-06-20): base was 4.50 / 5.30 / 6.00. Forward earnings power is up ~12% / ~17% / ~24%.
Valuation cross-check at $164.89 ``:
| Multiple on base | vs prior dossier (at $167.34) | |
|---|---|---|
| FY2026 | 32.7x `` | was 37.2x |
| FY2027 | 26.6x `` | was 31.6x |
| FY2028 | 22.1x `` | was 27.9x |
| FY2027 bear | 32.5x `` | — |
This is the refresh's most important number. The stock is down 1.5% and forward earnings are up 17%, so NVT de-rated by roughly five turns on FY2027 without falling. The prior dossier said "the risk/reward at $167 is worse than it was at $90." That is still true in absolute terms, but it is meaningfully less true than it was seven weeks ago. The bear line is the discipline: at the bear FY2027 of $5.08 you are paying 32.5x today, and a de-rating to 18–20x on that number puts the stock at $91–$102 `` — a 38–45% drawdown that is entirely available.
Note: our model create was deliberately not run — no committed base case in an unattended sweep.
Bull case (strengthened on three legs, weakened on one).
nVent is the way to own the AI data hall at an industrial multiple while it earns a pure-play growth rate. Q2 makes four claims that were previously assertions into facts: (i) $2B of data-center revenue in 2026, more than double last year — the vertical is now sized, not implied; (ii) the margin expands as it scales — Systems Protection ROS +150bps on 70% growth, adjusted company ROS +110bps while absorbing >$50M of inflation; (iii) the non-AI floor is inflecting, not eroding — Electrical Connections organic +17.9%, broad-based, on genuine distributor sell-through, three times what the sell side modelled; (iv) it converts to cash — FCF +125%, DSO and DIO both improved through a 47% organic quarter, net leverage 1.2x against a 2.0–2.5x target.
Add the capacity story: three liquid-cooling expansions in three years, >400,000 sq ft, Blaine 1 opened in ~100 working days and is running ahead of plan, Blaine 2 lands H1-2027, and an autumn modular, hot-swappable liquid-cooling platform with an attached service capability. That last item is the specific answer to the "you'll get relegated to gray-space metal-bashing" bear — modularity plus service plus installed base is the content-per-rack defence, and the customer set has broadened from hyperscalers to neo-clouds, colos, integrators and distribution.
The contrarian-bull point the market is currently mispricing: the stock fell 12.9% into the print on a sector multiple reset, then rose 7.8% on its own numbers, and net-net has gone nowhere while FY2027 earnings power rose ~17%. NVT is being traded as a beta expression of "AI infrastructure" at a moment when its own idiosyncratic execution is separating from the theme. Five turns of de-rating happened without a single bad datapoint from the company.
Bear case (two risks unchanged, one materially worse, one newly retired).
Pre-mortem (18 months out, thesis broke). It is early 2028. The Q3-2026 order re-acceleration management promised arrived, but Q1-2027 orders went flat as hyperscalers guided 2027 capex conservatively; the $2.5B backlog shipped through 2027 and was replaced at a lower rate; Blaine 2 opened into softening demand and the +40% capex build became fixed-cost drag; FY2027 organic came in at +6% instead of +18%, adjusted EPS at ~$5.08 instead of $6.20; the multiple compressed from 27x to 18x, and the stock sat in the mid-$90s. Most likely actual cause: an AI-capex digestion pause colliding with newly built capacity — the same risk as June, but now with more operating leverage pointed the wrong way.
Are multiples too high? Less so than in June. 26.6x FY2027 base is defensible for 18%+ organic growth, 22% adjusted ROS, 1.2x leverage and a genuine second growth vertical in power utilities. The problem is not the base case; it is that the bear case is not priced at all — 32.5x on bear FY2027 with a $91–$102 de-rating floor. The asymmetry has improved from clearly unattractive to roughly balanced. It has not become attractive.
Where the bull case is weakest, after Q2.
Lens 14 (Management questions) — carried. See the previous dossier. Questions 1, 2, 3 and 5 were partially answered by this quarter ($2B data-center target; broadened customer set; backlog "around the right level"; the ex-data-center line accelerating to +17.9%). Four new ones to stack on top:
Every dossier we have written on nVent Electric, newest first.
The order line decelerated and the backlog shrank
A boring enclosures-and-connectors company that bought and sold its way into the AI data-center build
Covered in the Knowledge Base
Energy & Power
| Industry | Energy |
| Size | Public Company |
Where nVent Electric sits against the other names we cover on this beat. Each line is that company’s verdict, not a summary of it.
The subsidy thesis just got a second leg — a $0.38/W federal module price floor from the Aug 6 Section 232 proclamation now sits ABOVE First Solar's o…
Cash $1.7B
The franchise got stronger and the price got cheaper
Cash $1.2B
The two mechanical bear points got repaired
Cash $3.0B
A best-in-class, AI-demand-levered renewable platform run through an externally-managed structure that skims ~$223M/yr in base fees plus IDRs off the…
A $1.1B-cap company sitting on a $3.5B storage backlog and 83 GWh dev pipeline, but the equity is a levered residual claim junior to $2.2B non-recours…
Cash $1.4B