Sum-of-the-Parts Valuation in Excel: Build an SOTP Model (2026)

June 26, 2026 · VeloraAI Team

Financial Modeling Excel Data Analysis

A diversified conglomerate trading at a 22% discount to the sum of its segment values isn't a market anomaly — it's a setup. Activist investors made $14B targeting break-up candidates in 2024 alone, and every pitch started the same way: a sum-of-the-parts valuation in Excel that proved the whole was worth less than its pieces. Whether you're an equity research analyst writing a Buy thesis on GE Aerospace, an IB associate prepping a divestiture pitch, or a PE deal team underwriting a carve-out, SOTP is the model that turns a one-line market cap into a defensible break-up thesis.

This guide walks through a full SOTP build in Excel: segmenting the financials, picking the right multiple for each business, allocating corporate overhead, bridging to equity value, and stress-testing the conglomerate discount. Every formula here works in Excel 365 and Excel 2021. Copy them into your model and ship a banker-grade output today.

What Is Sum-of-the-Parts Valuation?

Sum-of-the-parts (SOTP), also called "break-up" analysis, values a multi-segment company by valuing each business segment separately — using the multiple or DCF most appropriate for that segment's industry — then summing the enterprise values, allocating corporate overhead, and bridging to equity value through net debt and minority interest. It's the standard valuation method for conglomerates, holding companies, and any business where blended multiples mask real value.

The math is simple. The judgment is everything. Picking 11x EBITDA for an aerospace segment and 6x EBITDA for a power segment isn't arithmetic — it's a defensible comp set, a normalization of segment EBITDA, and a credible overhead allocation. Get those three right and your SOTP will survive an investment committee. Get them wrong and you've built a calculator that produces whatever number you want.

When SOTP Beats Other Valuation Methods

Use SOTP when a single multiple or single DCF can't reflect the underlying mix of businesses. The classic cases:

Industrial conglomerates (Honeywell, Siemens, 3M, Emerson) — aerospace deserves a different multiple than HVAC
Tech holding companies (Alphabet, Meta, Amazon) — cloud, ads, and hardware trade on different metrics
Media and entertainment (Disney, Warner Bros. Discovery) — streaming, parks, studios, and networks have distinct risk profiles
Activist break-up theses — quantifying spin-off uplift requires segment-level value
Carve-out and divestiture pitches — the seller needs a defensible standalone value per segment

If the company runs one business with one set of economics, don't waste time on SOTP. A DCF or trading comp does the job.

ℹ️ Note: SOTP is the only valuation method that explicitly quantifies the conglomerate discount — the gap between the standalone value of segments and the market cap of the combined entity. That gap is where activists make money.

How Do You Build a Sum-of-the-Parts Valuation in Excel?

To build an SOTP valuation in Excel, segment the company's revenue and EBITDA by business unit, select an appropriate trading or transaction multiple for each segment based on pure-play comps, calculate each segment's enterprise value, sum the segment EVs, subtract capitalized corporate overhead, then bridge from total EV to equity value by deducting net debt and minority interest and adding cash and non-operating assets.

Here's the step-by-step build that holds up to senior review:

Segment the financials. Pull segment revenue and EBITDA from the latest 10-K and reconcile to consolidated.
Normalize segment EBITDA. Strip non-recurring items, allocate stock comp, restore intersegment eliminations.
Build the comp set per segment. Pure-play comps only — no conglomerates valuing conglomerates.
Choose the segment multiple. EV/EBITDA, EV/Revenue, or DCF — match the metric to the industry.
Calculate segment EV. Multiple × segment metric = segment EV.
Allocate corporate overhead. Capitalize unallocated overhead at a discount rate and subtract from EV.
Bridge to equity value. EV minus net debt minus minority interest plus cash and non-operating assets.
Calculate implied share price and conglomerate discount. Compare to market cap.

Step 1: Segment the Financials

Open the latest 10-K and find the segment reporting footnote (usually Note 17 or Note 18). U.S. GAAP under ASC 280 and IFRS 8 require segment disclosure for operating segments. You want segment revenue, segment EBITDA (or operating profit + D&A), and segment assets if available.

The setup in Excel:

A                   B           C           D           E           F
Segment             Revenue     EBITDA      Margin      % Rev       % EBITDA
Aerospace           14,200      3,550       25.0%       =B3/B$7     =C3/C$7
Power Systems       8,400       1,260       15.0%       =B4/B$7     =C4/C$7
Industrial Auto.    6,100       1,098       18.0%       =B5/B$7     =C5/C$7
Software            2,800       980         35.0%       =B6/B$7     =C6/C$7
Consolidated        31,500      6,888       21.9%       100%        100%

⚠️ Warning: Segment EBITDA reported in the 10-K is almost never directly usable. Companies allocate corporate overhead inconsistently — some push it down to segments, others hold it in a "Corporate & Other" line. Always read the segment footnote in full and rebuild segment EBITDA to a consistent definition before applying multiples.

Normalize Segment EBITDA

Before you multiply anything, clean the EBITDA. The four normalizations that change every SOTP:

Stock-based compensation. If the company reports "Adjusted EBITDA" that excludes SBC, add it back. Investors pay for real cash earnings, not adjusted ones.
Restructuring and impairments. Strip out one-time charges that won't recur.
Intersegment eliminations. If Segment A sells to Segment B, eliminate the revenue and matching cost at the consolidated level — but track it at the segment level for standalone analysis.
Pension and OPEB. Service cost only at segment level; interest cost and amortization belong below the line.

Use a clean normalization schedule:

=SUMPRODUCT((Adjustments_Type<>"One-Time")*Adjustments_Amount*(Adjustments_Segment=A3))

This pulls all recurring adjustments tagged to the segment and excludes one-time items.

Step 2: Build the Segment Comp Set

This is where SOTP succeeds or fails. Pure-play comps only. If you're valuing GE's aerospace segment, your comp set is RTX, TransDigm, HEICO, and Howmet — not Honeywell or Lockheed (which have their own diversification).

The minimum comp set per segment:

Segment Type	Suggested Comps Count	Primary Multiple	Secondary Multiple
Mature industrial	5–8	EV/EBITDA	EV/Revenue
High-growth software	4–6	EV/Revenue	EV/ARR
Capital-intensive (energy, utilities)	6–10	EV/EBITDA	P/E
Cyclical commodity	8–12	EV/EBITDA (mid-cycle)	EV/Capacity
Subscription / SaaS	5–7	EV/ARR	EV/Revenue
Distribution / logistics	5–8	EV/EBITDA	EV/EBIT

Build the comp set in a separate sheet and pull the median multiple into the SOTP page with a structured reference:

=MEDIAN(IF(CompTable[Segment]=A3, CompTable[EV_EBITDA]))

In Excel 365, use the cleaner spilled-array version:

=MEDIAN(FILTER(CompTable[EV_EBITDA], CompTable[Segment]=A3))

This dynamically pulls the median EV/EBITDA for the comp set tagged to the segment in A3.

Step 3: Choose the Right Multiple Per Segment

Don't default to EV/EBITDA for every segment. Match the metric to how investors value the industry:

EV/EBITDA — mature, asset-heavy, positive-earnings businesses
EV/Revenue — high-growth or pre-profit segments where margin will scale
EV/EBIT — capital-intensive businesses where D&A distorts EBITDA comparability
P/E — regulated utilities, financial services, REITs (operating-income-based comps don't translate)
EV/ARR — pure SaaS or subscription segments
DCF — segments with predictable, long-duration cash flows (infrastructure, royalty streams)

💡 Pro Tip: For any segment that contributes more than 25% of total enterprise value, build a parallel DCF as a sanity check on the multiple-based output. If the DCF lands within ±15% of the multiple-based value, you have a defensible number. If the gap is wider, dig in — one of them is wrong.

The Two-Multiple Range Per Segment

Don't pick one multiple. Pick a range — typically the 25th and 75th percentile of the comp set — and present low/mid/high cases. This gives the model honest precision and is what every senior banker expects.

Low_EV  = Segment_EBITDA * PERCENTILE.INC(Comps_Range, 0.25)
Mid_EV  = Segment_EBITDA * MEDIAN(Comps_Range)
High_EV = Segment_EBITDA * PERCENTILE.INC(Comps_Range, 0.75)

Step 4: Calculate Segment Enterprise Value

With normalized EBITDA and chosen multiples, the segment EV calculation is straightforward. The structure:

A              B           C           D           E           F           G
Segment        EBITDA      Multiple    Method      Low EV      Mid EV      High EV
Aerospace      3,550       13.5x       EV/EBITDA   =B3*11.0    =B3*C3      =B3*16.0
Power          1,260       7.0x        EV/EBITDA   =B4*5.5     =B4*C4      =B4*8.5
Industrial     1,098       11.0x       EV/EBITDA   =B5*9.0     =B5*C5      =B5*13.0
Software       980         18.5x       EV/EBITDA   =B6*15.0    =B6*C6      =B6*22.0
Subtotal       6,888                               =SUM(E3:E6) =SUM(F3:F6) =SUM(G3:G6)

graph TD
    A[Segment Revenue & EBITDA] --> B[Normalize Segment EBITDA]
    B --> C[Select Pure-Play Comps]
    C --> D[Apply Multiple Range]
    D --> E[Segment Enterprise Value]
    E --> F[Sum Segment EVs]
    F --> G[Subtract Corporate Overhead]
    G --> H[Total Enterprise Value]
    H --> I[Less: Net Debt]
    H --> J[Less: Minority Interest]
    H --> K[Plus: Cash & Non-Op Assets]
    I --> L[Equity Value]
    J --> L
    K --> L
    L --> M[Implied Share Price]
    M --> N[Compare to Market Cap]
    N --> O[Conglomerate Discount]

Step 5: Allocate Corporate Overhead

This is the step most SOTP models get wrong, and it's the single biggest source of senior pushback. Every conglomerate has unallocated corporate overhead — CEO comp, board fees, headquarters real estate, central treasury, central IT. If you sum segment EVs without adjusting for this, you've overvalued the company.

There are three accepted methods. Pick one and apply it consistently:

Method	How It Works	When to Use
Allocate to segments	Push overhead down to segments pro rata (by revenue, EBIT, or assets) before applying multiples	When overhead is small (<5% of segment EBITDA) and easy to push down
Capitalize and subtract	Capitalize annual unallocated overhead at a 5–8x multiple, subtract as a negative line item	When overhead is material and you want it visible on the page
Mid-cycle treatment	Capitalize at the weighted average segment multiple	When overhead funds genuine cross-segment synergies (rare)

Method 1: Pro Rata Allocation

Segment_Overhead_Alloc = Total_Corp_Overhead * (Segment_Revenue / Total_Revenue)
Adjusted_EBITDA = Reported_EBITDA - Segment_Overhead_Alloc

Allocation by EBIT is more conservative since it weights toward higher-margin segments (which arguably benefit more from corporate functions). Allocation by revenue is the bank-default.

Method 2: Capitalize and Subtract

Corp_Overhead_Adj = -Annual_Corp_Overhead * Overhead_Multiple

Where Overhead_Multiple is typically 5–8x — below the segment-weighted average because corporate overhead is a cost stream, not a growth asset. Subtract this as a separate line item on the SOTP build so the deduction is visible to anyone reading the page.

⚠️ Warning: Senior reviewers will challenge you on the overhead multiple. Defend it with a real-world spin-off comp — when ABB spun off its Power Grids business to Hitachi, the stranded-cost overhead was capitalized at roughly 6x. Cite the precedent.

Step 6: Bridge from Enterprise Value to Equity Value

You've summed segment EVs and subtracted corporate overhead. The next bridge is mechanical but full of traps:

SOTP Enterprise Value           93,400
Less: Capitalized Overhead      (3,200)
Adjusted Enterprise Value       90,200

Less: Total Debt                (14,500)
Less: Pension Underfunding      (1,200)
Less: Minority Interest          (850)
Plus: Cash & Equivalents         8,400
Plus: Investments (at FV)        2,100
Plus: NOLs (tax-affected)        1,800
Equity Value                    85,950

Diluted Shares Outstanding       1,275
Implied Share Price             $67.41
Current Share Price             $54.20
Upside to SOTP                   24.4%

The bridge items that trip junior analysts up:

Minority interest at fair value — book value is rarely correct. If the minority interest sits in a high-growth segment, mark it up to a proxy of that segment's multiple.
Pension underfunding — net of tax effect. Use the projected benefit obligation minus plan assets, then tax-affect at the marginal rate.
NOLs — only credit value if the company has a clear path to using them. A net-loss conglomerate with a $5B NOL but no path to taxable income gets $0 credit.
Equity investments and minority stakes — value at the look-through fair value (apply the investee's multiple), not at book.

=SUMPRODUCT(InvestmentTable[Stake]*InvestmentTable[Investee_EV])

Step 7: Calculate the Conglomerate Discount

The conglomerate discount is the gap between standalone segment value and market value of the combined entity. The formula:

=(SOTP_Equity_Value - Market_Cap) / SOTP_Equity_Value

A 10–25% conglomerate discount is normal for diversified conglomerates. A discount above 30% signals a real break-up opportunity — and is the pitch every activist makes. A discount below 5% means the market already prices the segments correctly; SOTP isn't where you'll find alpha.

Example: If your SOTP equity value is $86B and the company trades at a $69B market cap, the conglomerate discount is ($86B − $69B) / $86B = 19.8%. That's the percentage uplift available if the conglomerate were broken up at standalone segment multiples — before factoring in friction costs, stranded overhead, and tax leakage.

Stress-Testing the Conglomerate Discount

A 20% discount can mean the market is wrong, or it can mean your multiples are too rich. Pressure-test both directions:

graph LR
    A[Observed Discount 19.8%] --> B{Sanity Checks}
    B --> C[Use 25th percentile multiples]
    B --> D[Add stranded overhead]
    B --> E[Apply tax leakage on separation]
    C --> F[Adjusted Discount]
    D --> F
    E --> F
    F --> G{Discount Still >10%?}
    G -->|Yes| H[Real Break-Up Opportunity]
    G -->|No| I[Market Pricing Is Fair]

The two most common reasons your discount is too high:

You ignored stranded overhead. Splitting a conglomerate doesn't eliminate the CEO, board, or shared services — those costs follow the segments and reduce standalone value.
You ignored tax leakage. Spin-offs above the Section 355 threshold require tax-free treatment; outside that, separations create real tax cost.

Build a sensitivity table that flexes (a) the segment multiples to the 25th percentile and (b) capitalized overhead 50% higher. If the discount is still above 10%, the thesis holds.

A Full Worked Example: A Diversified Industrial Conglomerate

Walk through a real-world structure. Assume a public industrial conglomerate ("IndustrialCo") with four segments:

Segment	LTM Revenue ($M)	LTM EBITDA ($M)	EBITDA Margin
Aerospace & Defense	14,200	3,550	25.0%
Power Systems	8,400	1,260	15.0%
Industrial Automation	6,100	1,098	18.0%
Industrial Software	2,800	980	35.0%
Corporate & Other	(1,200)	(320)	n/a
Consolidated	30,300	6,568	21.7%

Apply pure-play comp medians:

Segment	Median EV/EBITDA	Segment EV ($M)
Aerospace & Defense	13.5x	47,925
Power Systems	7.0x	8,820
Industrial Automation	11.0x	12,078
Industrial Software	18.5x	18,130
Sum of Segments		86,953

Apply overhead and equity bridge:

Sum of Segment EV                86,953
Less: Capitalized Corp Overhead  ( 2,240)    ($320M annual × 7x)
Adjusted Enterprise Value        84,713

Less: Total Debt                 (14,500)
Less: Pension Underfunding       ( 1,200)
Less: Minority Interest          (   850)
Plus: Cash & Equivalents          8,400
Plus: Equity Investments          2,100
Plus: NOLs (tax-affected)         1,800
Equity Value                     80,463

Diluted Shares Outstanding        1,275
Implied SOTP Share Price         $63.11
Current Share Price              $52.50
Conglomerate Discount             16.8%

The 16.8% discount is in the normal range for industrials. The break-up thesis lives or dies on whether you can defend the Aerospace multiple at 13.5x and the Software multiple at 18.5x with credible pure-play comps.

Common Pitfalls That Sink SOTP Models

After auditing hundreds of SOTP models, the same five mistakes show up over and over:

1. Comp Pollution

Putting Honeywell in your comp set when valuing GE's aerospace segment is comp pollution. Honeywell is itself a conglomerate — its multiple already reflects a conglomerate discount. You need true pure-plays.

2. Double-Counting Synergies

If you allocate corporate overhead to segments AND deduct capitalized overhead at the EV level, you've subtracted it twice. Pick one method per model and stick with it.

3. Using Reported Segment EBITDA Without Normalization

Companies game segment reporting to flatter favored businesses. Always rebuild segment EBITDA from segment revenue minus segment direct costs, with a documented overhead allocation.

4. Ignoring Tax Leakage on Spin-Offs

A standalone segment value of $50B doesn't mean $50B reaches shareholders. Section 355 requires specific structural conditions; outside those, separations trigger taxable gain at the corporate level. Build a 10–15% tax friction assumption into the break-up case.

5. Stale Comps

A median EV/EBITDA from December 2024 won't survive a 2026 review. Refresh comps within 30 days of any pitch. Use Power Query or a Capital IQ pull, not a static copy-paste.

💡 Pro Tip: Tag every comp with its data date in a hidden column. When the SOTP is reopened, a quick =MAX(CompTable[DataDate]) tells you whether the analysis needs a refresh before you can present it.

How Do You Defend an SOTP in an Investment Committee?

To defend an SOTP in an IC, lead with the segmentation rationale (why these four segments, not three or five), then the comp methodology (pure-play, comp date, percentile range), then the corporate overhead treatment (allocation method and capitalization multiple), and finally the equity bridge (every line item with a source). A two-way sensitivity analysis in Excel across segment EBITDA multiples and overhead capitalization rates turns the static SOTP into a range that survives IC scrutiny — show 25th/75th percentile multiples alongside the median so the committee can see the downside case immediately. Sensitivity tables on segment multiples and overhead drive the answers to the questions the committee will actually ask.

The three questions every senior partner asks:

"What's your worst case?" — Show the 25th percentile multiple SOTP. If it still produces a discount, the thesis is robust.
"What's the friction cost of separation?" — Quantify stranded overhead, tax leakage, dis-synergies, and one-time separation costs as a percentage of segment EV.
"What's the comparable spin-off precedent?" — Reference a real-world spin (United Technologies → Otis + Carrier, or DowDuPont → DuPont + Corteva + Dow). Cite the discount that closed at separation.

Linking SOTP to Other Valuation Methods

SOTP sits inside a broader valuation toolkit. Use it alongside, not instead of, the standard methods:

DCF — sanity-check segment multiples on the two largest segments with standalone DCFs.
Trading comps — comps you'd build for a single business become segment comps in an SOTP.
Precedent transactions — overlay control premiums on segment EVs to estimate take-out value per segment.
Football field — display SOTP, DCF, comps, and precedents side-by-side. See our football field valuation chart guide for the visual.

If you're new to building the underlying multiples, start with our comparable company analysis guide and precedent transactions analysis guide — both feed directly into the SOTP comp tables. For M&A situations where the acquirer is evaluating a conglomerate acquisition, pair the SOTP with a merger model in Excel to test whether buying the whole entity is accretive relative to a targeted segment acquisition.

Frequently Asked Questions

What's the difference between SOTP and DCF valuation?

A DCF values a company as a single cash flow stream discounted at one WACC. SOTP values each segment separately, often using different multiples per segment, then sums them. Use SOTP when segments have materially different risk, growth, or capital intensity — like an industrial conglomerate. Use DCF when the company is fundamentally one business.

How do you handle segment-level capex and working capital in SOTP?

If you're applying EBITDA multiples, segment capex and working capital are embedded in the multiple — the comp set already trades reflecting its capex intensity. If you're building a segment-level DCF, pull segment-level capex from the 10-K (usually disclosed by segment) and estimate segment working capital from segment receivables and inventory if disclosed; otherwise allocate consolidated working capital pro rata by segment revenue.

What multiple should I use for corporate overhead?

Capitalize annual corporate overhead at 5–8x, below the segment-weighted average multiple. Overhead is a cost stream with no growth, so it deserves a discount-to-segment-average multiple. For most large-cap industrials, 6–7x is the bank-default. Always document why you picked the specific multiple — auditors and IC members will ask.

Should SOTP equity value match the market cap?

Almost never. If they match, the market already prices the segments correctly and SOTP isn't actionable. A discount of 10–25% to market is normal for conglomerates and signals the standard conglomerate discount. A discount above 30% suggests a real break-up opportunity. A premium (SOTP < market cap) usually means your multiples are too low or you've missed an asset on the equity bridge.

Can I use SOTP for a private company?

Yes — and it's often the cleanest method for private holding companies, family offices, and private equity portfolios. Replace traded comps with M&A transaction multiples for each segment's industry. Apply an additional 15–25% private-company discount (illiquidity, smaller deal size, lower disclosure) to the segment EVs. SOTP also drives LP reporting in PE funds: each portfolio company is, in effect, a segment in the fund-level SOTP.

The Bottom Line

A sum-of-the-parts valuation is the analyst's sharpest tool for cutting through conglomerate complexity. The math is trivial. The judgment — segmentation, pure-play comps, overhead allocation, equity bridge — is where the work happens. Build it disciplined, document every multiple, and the SOTP will survive every investment committee in the room.

If you're building SOTPs across multiple companies or refreshing comp medians monthly, VeloraAI can pull pure-play comp sets from a natural-language segment description, normalize segment EBITDA from 10-K disclosures, and reconcile the equity bridge against your filings so the model ties out before the deck goes to the partner. For the next step, pair this with our DCF model guide for the segment-level cross-check, and our WACC calculation guide for the discount rates that anchor any segment DCF.