Industrial Conglomerates and Sum-of-the-Parts Valuation: GE, Honeywell, 3M, and Danaher

The Conglomerate Discount Is Real — and Persistent

Wall Street has a conglomerate problem. Diversified industrial companies trade at 10–25% below their sum-of-the-parts value, a discount that has persisted across market cycles for over three decades. Academic research from LeBaron and Speidell (1987) first documented the phenomenon, and subsequent studies have consistently confirmed it: the median US conglomerate trades at a 12–15% discount to its disaggregated SOTP value.

The discount exists for rational reasons. Capital allocation opacity, management bandwidth constraints, investor self-selection preferences, and compensation misalignment all contribute. But rational does not mean permanent. When catalysts emerge — activist campaigns, new management teams, or strategic pivots — that discount can close rapidly, generating 20–40% returns in 12–24 months. The trick is identifying which conglomerates are ripe for change and which will remain value traps indefinitely.

GE proved the thesis in the most dramatic fashion possible. The rest of the industrial complex is paying attention.

How to Build an SOTP Model: Segment-by-Segment Valuation

The SOTP methodology is straightforward in concept but demanding in execution. You value each business segment using comparable pure-play multiples, subtract net debt and corporate overhead, and compare the result to the current enterprise value. The gap is your opportunity — or your trap.

Step 1: Segment Identification and Revenue Allocation

Start with the 10-K segment disclosures. Companies report revenue, operating income, depreciation, and sometimes capital expenditures by segment. Honeywell, for instance, reports four segments: Aerospace Technologies ($15.5B revenue), Industrial Automation ($6.7B), Building Automation ($6.1B), and Energy & Sustainability Solutions ($5.8B). Each segment has distinct growth profiles, margin structures, and comparable company sets.

Step 2: Comparable Selection and Multiple Application

This is where the art enters. Honeywell Aerospace competes with GE Aerospace, RTX, and Safran — pure-play aerospace companies trading at 28–35x NTM earnings. Honeywell's automation businesses compare to Emerson, Rockwell, and Siemens' digital industries at 20–25x. The blended Honeywell multiple of ~22x understates the value of the aerospace segment (60%+ of EBIT) while overstating the value of the slower-growth industrial segments. This is exactly where value hides.

For a deeper dive into selecting appropriate valuation multiples, see our relative valuation guide.

Step 3: Corporate Overhead and Stranded Costs

The most common SOTP error is ignoring corporate costs. A diversified industrial typically carries $500M–$1.5B in annual corporate overhead that would need to be allocated or eliminated in a breakup. Not all of it disappears. Shared services for IT, legal, HR, and treasury must be replicated at each standalone entity. We typically assume 40–60% of corporate overhead is stranded for the first 2–3 years post-separation, reducing SOTP value by 3–7%. Analysts who ignore this overstate the upside.

Industrial Conglomerate SOTP Analysis

The GE Playbook: Why Breakups Work

GE's transformation from a $60B market cap conglomerate in 2018 to three focused entities collectively worth north of $350B is the case study that reshaped how the market thinks about industrial diversification. Larry Culp executed the playbook that Jack Welch's successors should have run a decade earlier.

The logic was simple. GE Aerospace — with LEAP engines powering every Boeing 737 MAX and Airbus A320neo, plus a $150B+ services backlog — deserved a 30x+ earnings multiple as a pure-play aerospace OEM and aftermarket services business. But the market was applying a blended 15–18x multiple to all of GE because the power and healthcare divisions dragged down the growth narrative. GE HealthCare, separated in January 2023, now trades at 20–22x earnings as a medtech pure-play. GE Vernova captured the energy transition theme as a standalone entity with improving renewable energy margins.

The value creation was not magical. It was mathematical. Each entity found its natural investor base, optimized its capital structure, and received a multiple appropriate to its growth and margin profile. The conglomerate discount evaporated the moment the conglomerate ceased to exist.

Honeywell: The Next Breakup Catalyst

Elliott Management disclosed a significant stake in Honeywell in late 2024 and publicly advocated for a separation into aerospace-focused and automation-focused entities. The thesis is compelling. Honeywell Aerospace Technologies generates roughly $15.5B in revenue with segment margins near 28%, competing with pure-play aerospace companies trading at 30–35x NTM earnings. The remaining automation and buildings businesses, growing at mid-single-digits with 20–23% margins, compare to industrial automation peers at 20–25x.

Management initially pushed back. CEO Vimal Kapur argued that the diversified portfolio provides resilience and cross-selling opportunities. This is the standard conglomerate defense — and it almost never holds up under scrutiny. Cross-selling synergies between aerospace parts and building management systems are minimal. The “resilience” argument is another way of saying that underperforming segments dilute the multiple of outperforming ones.

In early 2025, Honeywell announced plans to separate into three standalone companies — Honeywell Automation, Honeywell Aerospace, and a spun-off Advanced Materials division. Elliott's pressure worked. The market responded favorably, but the full value creation will depend on execution, separation costs, and the multiples each entity commands once trading independently.

Understanding how management teams allocate capital across segments is central to the conglomerate thesis — we explore this in depth in our capital allocation analysis guide.

Danaher vs. Siemens: Two Models for the Multi-Industrial

Danaher and Siemens represent opposite ends of the conglomerate management spectrum — and the returns tell the story.

Danaher's model is acquire, improve, and prune. The Danaher Business System (DBS) is a proprietary lean operating methodology applied to every acquisition, targeting 200–400 basis points of margin expansion within 3–5 years. The company has spun off Fortive (2016), Envista (2019), and Veralto (2023), demonstrating willingness to separate businesses when the portfolio logic no longer holds. The result: Danaher has compounded at roughly 16% annually since 2003, versus approximately 10% for the S&P 500. The stock trades at 25–30x NTM earnings — essentially no conglomerate discount.

Siemens, by contrast, spent decades as a sprawling European conglomerate with exposure to everything from gas turbines to hearing aids. The company has progressively simplified: the Siemens Energy spin-off (2020), Siemens Healthineers IPO (2018), and Innomotics sale (2024) have narrowed the focus toward digital industries and smart infrastructure. Siemens now trades at 18–22x NTM earnings, a premium to most European industrials but still below US peers. The discount partly reflects European market dynamics and partly reflects the remaining portfolio complexity.

The lesson: active portfolio management eliminates the conglomerate discount. Passive portfolio management perpetuates it. Investors should track management's willingness to divest as a key indicator of future value creation.

3M and Emerson: Turnaround Bets with SOTP Upside

3M is the contrarian bet in this group. After years of litigation overhang (PFAS and combat earplugs), margin compression, and organic growth deceleration, the stock traded down to roughly 11–12x NTM earnings by mid-2024, a 40% discount to its 10-year average multiple. The Solventum healthcare spin-off completed in April 2024 removed $8.5B of healthcare revenue and a significant portion of the litigation risk. What remains is a $25B industrial company with strong brands in safety, adhesives, and abrasives, margins depressed at 19–20% with a path back to 22–24% under new CEO Bill Brown.

The math works if management executes. At $6B in EBITDA (assuming margin recovery) and an 18x multiple (reasonable for a mid-teen ROIC industrial), 3M's equity is worth $100–110B versus roughly $75B today. That is 30–40% upside with a 3% dividend yield while you wait. The risk is that margin recovery stalls and organic growth remains below 2%, justifying the current discount.

Emerson's story is different but equally interesting. CEO Lal Karsanbhai has transformed Emerson from a diversified industrial into a focused automation and software company, acquiring a majority stake in AspenTech and divesting the climate technologies business to Blackstone. The remaining portfolio — process automation, discrete automation, and industrial software — generates 25–27% EBITDA margins with a path toward 30%+ as the software mix increases. Full ownership of AspenTech, which the market expects by 2026, would create a pure-play automation and software platform warranting 22–25x NTM EBITDA.

When Breakups Destroy Value: The Cautionary Cases

Not every breakup creates value. The market learned this the hard way with several notable failures. DowDuPont's three-way split into Dow, DuPont, and Corteva created three companies that collectively underperformed the S&P 500 in the years following separation. The issue: commodity chemicals (Dow) and agriculture (Corteva) faced cyclical headwinds that offset the SOTP re-rating, and the promised $3.5B in cost synergies from the original merger proved difficult to replicate as standalone entities.

The conditions that favor value-creating breakups are specific: (1) segments with genuinely different growth profiles and investor bases, (2) minimal genuine operational synergies between segments, (3) each standalone entity having sufficient scale to operate independently, and (4) management teams with deep domain expertise ready to lead each entity. When these conditions are absent — when the synergies are real, or the standalone entities are too small to attract institutional ownership — the breakup simply creates two or three worse companies instead of one mediocre one.

Frequently Asked Questions