Sum of the Parts (SOTP) Valuation: What It Is, How to Build One, and How to Answer It in Interviews

Some companies are really three or four different businesses sharing one stock ticker. Valuing them with a single multiple is like pricing a fruit basket by the pound when half of it is truffles.

That's where sum of the parts (SOTP) valuation comes in. You value each business separately, add them up, then bridge to a share price. It shows up in pitch books, activist campaigns, spin-off analyses, and interviews.

This post walks through the logic, a full worked example with numbers, and the questions interviewers actually like to ask.

What Is a Sum of the Parts Valuation?

A sum of the parts valuation values each segment of a company on its own terms, using the methods and multiples that fit that specific business. You then add those values together and adjust for things that don't belong to any one segment.

The core idea: a slow-growing industrial division and a high-growth software division shouldn't get the same multiple just because they share a parent company.

When Bankers Use SOTP

You'll reach for SOTP when:

• The company is a conglomerate with segments in different industries (industrials, healthcare, consumer, etc.)
• Segments have very different growth rates or margins, so a blended multiple hides a lot
• A spin-off, carve-out, or divestiture is on the table, and you want to know what the pieces might be worth standalone
• An activist investor argues the company is undervalued and should break itself up
• The company holds stakes in other businesses, like a holding company with listed and unlisted investments

If a company does one thing (say, a pure-play software firm), SOTP doesn't add much. A normal comps or DCF analysis is enough.

How to Build a SOTP Valuation, Step by Step

Step 1: Break the Company Into Segments

Start with the reported segments in the company's filings. Public companies disclose segment revenue and some measure of segment profit, which is usually your raw material.

Sometimes you'll need to make judgment calls, like splitting a segment further if research analysts or management disclose more detail.

Step 2: Pick the Right Method for Each Segment

This is where SOTP earns its keep. Each segment gets the valuation approach that suits it:

• EV/EBITDA for most operating businesses (industrials, consumer, software)
• EV/Revenue for high-growth segments that aren't yet profitable
• P/E or P/Book Value for financial services arms (more on this trap later)
• DCF for segments with unique cash flow profiles, if you have enough data
• Market value for stakes in listed companies, often with a discount for tax or liquidity

Step 3: Choose Segment-Specific Comps

The software segment should be compared to software companies, not to the parent's industrial peers. Each segment gets its own peer group and its own multiple.

Keep your time period consistent. If you use NTM EBITDA for one segment, use NTM for all of them.

Step 4: Add Up the Segment Values

Multiply each segment's metric by its multiple and sum them. That gives you the gross enterprise value of the operating segments.

Step 5: Adjust for Everything Outside the Segments

This is the step people forget. You still need to account for:

• Unallocated corporate costs (head office, shared IT, the CEO's salary)
• Non-operating assets like equity stakes in joint ventures
• Net debt, which usually sits at the parent level
• Non-controlling interests and other debt-like items (pension deficits, preferred stock)

Step 6: Get to Equity Value Per Share

Enterprise value plus non-operating assets, minus net debt and other claims, gives equity value. Divide by diluted shares outstanding and you have your implied share price.

A Worked SOTP Example

Let's value a fictional company, Northwind Holdings, with three segments. Figures are NTM and in $ millions.

The Segments

Notice the software segment is under 20% of EBITDA but almost a third of the value. That's exactly the kind of thing a blended multiple would bury.

Handling Corporate Costs

Northwind has $50 million of unallocated corporate costs each year. Those costs are real and ongoing, so they reduce value.

A common approach is to capitalize them at the blended multiple of the segments: $9,600 / $800 = 12.0x.

• Corporate cost drag: $50 × 12.0x = ($600)
• Adjusted enterprise value: $9,600 − $600 = $9,000

Bridging to Equity Value

Now assume Northwind also has:

• A minority stake in a joint venture worth $300 (not consolidated, so not in EBITDA)
• Net debt of $2,500
• Non-controlling interests of $200
• 200 million diluted shares

Comparing to the Market

Say Northwind trades at $27.00 per share. That's a market equity value of $5,400 million.

• Gap: $6,600 − $5,400 = $1,200
• Implied discount to SOTP: $1,200 / $6,600 = about 18%

Run the math the other way and the market is valuing the whole company at roughly 10.4x its net EBITDA ($5,400 + $2,500 + $200 − $300 = $7,800 enterprise value, divided by $750 of EBITDA after corporate costs). That's well below the 12.0x blended segment multiple.

This gap is called the conglomerate discount, and it's the whole reason activists love SOTP slides.

Why Do Conglomerates Trade at a Discount?

Interviewers love this one. Good reasons to mention:

• Capital allocation concerns: investors worry cash from strong segments gets poured into weak ones
• Complexity: harder to analyze means fewer analysts and investors fully understand it
• Investor preference: many funds want pure-play exposure and can diversify on their own
• Management focus: leadership attention gets spread across unrelated businesses
• Lack of transparency: segment disclosure is often thinner than a standalone company's

Some practitioners apply an explicit discount to their SOTP value. Others leave it out and treat the gap as the "opportunity." Either way, know that the discount exists and why.

The Financial Services Trap

Here's a question that separates prepared candidates from everyone else.

Suppose a manufacturer owns a captive finance arm that lends to customers. That arm carries a lot of debt, but for a lender, debt is basically raw material, not financing.

So you'd typically:

• Value the finance arm on an equity basis (P/E or P/Book Value)
• Add that equity value in the bridge rather than mixing it into enterprise value
• Make sure the finance arm's debt is not subtracted again with the parent's net debt

Mixing enterprise value and equity value methods without adjusting is one of the most common SOTP errors.

If the Parts Are Worth More, Why Not Just Break Up?

Great follow-up question, and the honest answer is that breakups have real costs:

• Dis-synergies: shared procurement, distribution, or tech gets more expensive separately
• Stranded costs: some corporate overhead doesn't disappear when a segment leaves
• Tax leakage: selling a segment can trigger taxes (spin-offs can sometimes be structured to be tax-free, but not always)
• Transaction costs: advisory fees, legal work, new boards, new systems
• Smaller scale: standalone businesses might get worse borrowing terms or lose index inclusion

A strong answer acknowledges that SOTP shows the theoretical value, and the realistic value of a breakup is that number minus these frictions.

Common SOTP Mistakes to Avoid

• Forgetting corporate costs. Skipping them inflates the valuation, sometimes significantly.
• Double counting debt or cash. Be clear about what sits at segment level versus parent level.
• Mixing time periods. LTM for one segment and NTM for another makes the sum meaningless.
• Using the parent's peers for every segment. That defeats the entire purpose.
• Ignoring minority interests inside segments. If a consolidated subsidiary is 70% owned, the other 30% isn't yours.
• Treating the SOTP value as a guaranteed price target. It's a framework, not a promise.

How to Answer "Walk Me Through a Sum of the Parts" in an Interview

Keep it tight. Something like:

"A sum of the parts values each business segment separately, using the most appropriate method and peer multiples for that segment. You add the segment values to get a gross enterprise value, subtract capitalized corporate costs, add non-operating assets, and subtract net debt and other claims like non-controlling interests to get equity value. Divide by diluted shares for a per-share value. It's most useful for conglomerates or companies considering a spin-off, and comparing it to the market price often shows a conglomerate discount."

That's about 30 seconds and hits every key point. Then let the interviewer push you on corporate costs, the discount, or the financial services trap, because now you're ready for all three.

The Bottom Line

Sum of the parts is just comps or DCF done segment by segment, plus a careful bridge at the end. The concept is simple. The details (corporate costs, consistent periods, equity versus enterprise value) are where candidates slip up.

Practice the Northwind example on paper once or twice. If you can rebuild it from memory and explain why the market might not pay the full $33, you'll handle any SOTP question an interviewer throws at you.