Business

Schneider Electric — Know the Business

Figures converted from EUR at historical FX rates — see data/company.json.fx_rates. Ratios, margins, and multiples are unitless and unchanged.

Bottom line. Schneider is not an industrial machinery company; it is a regulated, channel-led electrical-infrastructure platform with a software and services flywheel bolted on, and the largest single corporate exposure to AI data-center power capex. The right way to read it: ~78% of revenue is Energy Management running at a 21.8% margin and growing double digits, and ~22% is Industrial Automation running at 14.2% margin with a true short cycle. The market is paying a 36x earnings multiple because it has — correctly — figured out the first half. The debate worth having is whether the data-center premium is durable as hyperscaler capex normalises, and whether the IA segment will ever close the margin gap to Rockwell.

1. How This Business Actually Works

The economic engine is one installed base monetised three ways. Schneider sells a code-certified piece of electrical hardware, then attaches a project / service contract for the next 10–30 years, then sells software that runs on top of it. The hardware is the foot-in-the-door; services and software are where the margin sits.

The single most important sentence in the operating model: incremental dollars of revenue from data centers and from software earn meaningfully more than incremental dollars of residential or discrete industrial revenue, and management has spent six years shifting the mix toward the high-margin ends. The 50bps of organic margin expansion in 2025 is not cost-cutting; it is mix. Bargaining power tracks scarcity: in 2022–24, lead times stretched and the OEM held it (Schneider raised price 4–6 points a year); by 2025 lead times normalised, so the next phase of margin expansion has to come from mix and software, not pricing.

The business is asset-light by industrial standards — capex 2.5–2.7% of sales, FCF conversion above 100% in 2025 — but it is not capital-light at the level of working capital. Inventory and receivables tied up in the backlog of switchgear and prefab modules absorb the first hit in any downturn, before margins fall.

2. The Playing Field

The cohort splits into three economic types and Schneider is the only one with meaningful exposure to all three: regulated electrical hardware (Legrand, Eaton, Schneider), industrial automation (Rockwell, Siemens DI, Schneider IA, ABB Robotics), and data-center electrical (Eaton, Schneider, Vertiv, ABB Electrification). The takeaway from the peer table is that Eaton has the highest margin in the cohort, ABB the highest ROIC, Rockwell the highest multiple — and Schneider sits second on each.

What the picture shows: Eaton, Legrand and Schneider cluster on margin (~19%) but the market pays Eaton and Rockwell a substantially higher multiple for the same dollar of EBITDA. The premium is not for margin — it is for perceived data-center purity. Eaton is the cleanest US-listed data-center power name; Rockwell is the cleanest US automation name. Schneider is bigger and arguably better positioned in data-center power than either, but is taxed for being European, blended with a slower IA segment, and held by a different shareholder base.

Siemens is the outlier and a useful benchmark for what Schneider has resisted becoming: a sprawling conglomerate (Mobility, Healthineers, Smart Infra, Digital Industries) where the group margin is dragged down by lower-quality segments. Schneider's choice to divest Industrial Automation hardware is not on the table — the bull case argues software pulls the IA margin up to the EM line.

3. Is This Business Cyclical?

Yes — but the cycle hits the two segments very differently, and a single "is electrical equipment cyclical" question hides the fact that Schneider is now a barbell of one structural-growth engine and one short-cycle engine.

Energy Management compounds. Its margin has expanded almost every year since 2019 and the data-center pull added a step-change in 2023–25. It is sensitive to residential construction at the bottom of the mix, but secure-power and grid demand has absorbed almost any soft patch. Industrial Automation tells the real cycle story: it grew through 2021–22, shrank in 2024 (mid-cycle pause that hit Rockwell and ABB Robotics the same way), and has just returned to +3% organic in 2025 with +4% in Q1 2026. The 2024 IA dip was a 7-percentage-point margin hit (17% → 15%) — a useful template for what a downturn looks like inside Schneider.

The structural protection in this cycle is the $29.8B backlog, up 18% in 2025, equal to roughly eight months of revenue and skewed toward data centers where contracts are increasingly multi-year, take-or-pay. That is not a number this business carried in 2008. The first place a downturn now shows is in residential wiring devices and in IA orders — both small enough to bend the group line by a few points without breaking it.

4. The Metrics That Actually Matter

P/E and revenue growth tell you almost nothing about whether the value engine is intact. Five numbers do.

ROIC has lifted 350 basis points in six years while revenue grew ~45%. That combination — margin up, returns up, scale up — is the textbook signature of mix shift, and it is rarer than a casual reader of industrial earnings would expect.

5. What Is This Business Worth?

The value is best understood as one economic engine with two clearly different parts, not a sum-of-parts spreadsheet. Energy Management is the engine — call it earnings power times reinvestment runway — and the data-center vertical is the option on top of it. Industrial Automation is a separate, lower-multiple business that the market currently bundles into the group at a discount.

A sanity-check using segment economics: applying a clean Eaton multiple to EM (28x EV/EBITDA × ~$8.3B EM EBITDA ≈ $235B EV) and a Rockwell-discount multiple to IA (~20x × ~$1.5B ≈ $30B EV) lands at roughly $265B EV — materially above the current ~$188B EV. The gap is the holding-company-style discount the European listing carries; closing it would require either US dual-listing, faster IA recovery, or a software ARR disclosure that lets investors value EcoStruxure separately. Treat the segment math as a measure of how much segment value the current multiple is not giving Schneider, not as a price target.

6. What I'd Tell a Young Analyst

Watch the segment-level book-to-bill split, not the group number — they move at different times in the cycle and the group line smooths over a turn in Industrial Automation. Treat data-center order share as a leading indicator of margin mix; once it crosses 35% the group margin floor structurally rises. The single most under-followed disclosure is software ARR (AVEVA, EcoStruxure, ETAP combined); a first clean break-out is the closest thing to a structural re-rating catalyst Schneider can produce, because the market today values the entire firm as an industrial. Do not anchor on Siemens — Schneider is more focused. Do anchor on Eaton — that is the multiple gap to argue with.

The thesis breaks if either (a) hyperscaler capex turns into a true bust rather than a digestion, (b) IA never gets to 17–18% margin, or (c) tariff or regulatory friction breaks the regional hub model. Everything else is noise.