Visit Website →
← Back to blog

Sum of the Parts Valuation: How SOTP Works, Step by Step

October 25, 2026
Sum of the Parts Valuation: How SOTP Works, Step by Step

Sum of the parts valuation (SOTP) estimates what a company is worth by valuing each business segment on its own and adding the results together. You then subtract net debt and other claims, and add non-operating assets, to get an equity value you can compare with the stock price. It is the right tool when one company contains very different businesses. It is a poor tool when the segments are really one business with a few labels on it.

Below you will find the formula, a worked example with round numbers, the mistakes that make SOTP numbers look better than they are, and a simple way to test the answer before you act on it.

Oracleinvestments
Compare Stocks Beyond One Valuation
Oracle Investments scores over 260 stocks on profitability, valuation, and financial health, with instant comparisons and real-time portfolio tracking.
Explore stock analysis

TL;DR

  • SOTP values each segment separately, usually with an EV/EBITDA multiple suited to that segment, then sums them into an enterprise value.
  • Equity value = sum of segment values, minus net debt and non-operating liabilities, plus non-operating assets.
  • It works best for conglomerates and holding companies. It is easy to bend: more segments, friendlier multiples and ignored overhead all push the answer up.
  • Cross-check the result with a DCF or owner earnings, and decide whether a conglomerate discount applies.
  • A high SOTP value is a hypothesis about the stock, not proof that it is cheap.

Table of Contents

What sum of the parts valuation actually measures

SOTP is also called break-up analysis, because it asks what the divisions would be worth if the company were split up. Wall Street Prep describes it as valuing each segment separately and adding them to reach an implied total enterprise value. Investopedia notes it is especially useful for conglomerates, where the value of individual units may exceed the value the market gives the whole.

The logic is simple. A single blended multiple applied to a company that owns a slow-growth industrial business and a fast-growing software business will misprice both. The industrial arm deserves a lower multiple, the software arm a higher one. Valuing them together hides that. Valuing them apart shows it.

That is also why the method is not useful for a focused company. If a business sells one product line to one type of customer, there is nothing to separate. A regular multiple or a discounted cash flow (DCF) will tell you more.

The SOTP formula

The structure is the same whichever multiple you use:

Equity value = sum of segment values − net debt − non-operating liabilities + non-operating assets

This is the formulation Investopedia lays out. Each piece has a job:

  • Segment values come from applying a multiple (or a DCF) to each segment's earnings or cash flow.
  • Net debt is total debt minus cash. Because segment values are enterprise values, debt must come out to reach equity. If this part is unfamiliar, our guide to enterprise value and how to calculate it covers it, and EV versus market cap explains why mixing the two produces bad comparisons.
  • Non-operating assets are things that earn nothing inside the segments, such as a minority stake in another company or excess real estate. They are added back because the segment earnings do not capture them.
  • Non-operating liabilities might include pension deficits or litigation reserves, depending on how you treat them.

Divide the resulting equity value by diluted shares outstanding and you have a per-share figure.

How to do a SOTP valuation in five steps

1. Identify the segments

Start with the company's annual report. U.S. companies disclose operating segments under ASC 280, and the segment note gives you revenue, operating profit and often assets by segment. Our piece on segment reporting analysis and the 10% and 75% tests explains how those disclosure thresholds work and what they leave out.

Stick to reported segments where you can. Research on analyst practice points out that analysts sometimes carve a company into more pieces than the filings show, which can build a story that the whole is far undervalued. If you split a segment further, be able to say why the pieces deserve different multiples.

2. Pick a valuation method for each segment

Most analysts reach for EBITDA multiples. A study of how analysts implement SOTP found that EBITDA multiples are the typical choice. That makes sense: EV/EBITDA is capital-structure neutral and easy to compare across peers. If you want a refresher, see EV to EBITDA as a valuation multiple.

EBITDA is not always the right metric. A capital-heavy business with large maintenance spending can look better on EBITDA than it deserves. For those segments, EV/EBIT or a segment-level DCF is more honest.

3. Choose comparable multiples

Find publicly traded companies that resemble each segment, not the parent. For a software segment, look at software peers. For a packaging segment, look at packaging peers. Use a range rather than a single number, and note whether the peers are bigger, faster-growing or more profitable than the segment you are valuing.

4. Allocate corporate costs

Headquarters costs belong to someone. Macabacus warns that leaving overhead unassigned makes segment valuations misleadingly high, and says experienced analysts spread those costs across segments, for example by revenue or EBIT share. The alternative is to capitalize the overhead as a negative line item, which is what the example below does.

5. Bridge to equity value and per-share value

Sum the segments, subtract net debt and other claims, add non-operating assets, and divide by diluted shares. Then decide whether to apply a discount, as covered below.

SOTP bridge from segment values to per-share value

A worked example with round numbers

The company below is hypothetical. It exists only to show the arithmetic.

Northfield Industries has three reported segments and 100 million diluted shares. The stock trades at $44.

SegmentEBITDAMultipleSegment value
Industrial equipment$400M9x$3,600M
Software and services$150M16x$2,400M
Consumer products$200M8x$1,600M
Sum of segments$7,600M
Unallocated corporate cost−$50M10x−$500M
Enterprise value$7,100M

Now the bridge:

  • Enterprise value: $7,100M
  • Less net debt: $1,800M
  • Plus non-operating assets (a minority stake): $200M
  • Equity value: $5,500M, or $55 per share

Against a $44 stock, that looks like 25% upside. But the analysis is not finished. Suppose you believe management allocates capital poorly across the divisions and apply a 15% conglomerate discount to enterprise value. EV falls to about $6,035M, equity value to about $4,435M, and the per-share figure to roughly $44. The apparent bargain disappears: the market price already reflects a discount of that size.

That is the real use of SOTP. It does not hand you a target. It shows what you must believe for the stock to be cheap.

The conglomerate discount: how much, if any?

A conglomerate discount is the idea that a multi-business company trades below the sum of its parts. The reasons are familiar: investors dislike opaque cross-subsidies, and management may spread capital across divisions instead of directing it to the best ones.

It is not a law of nature, though. A student valuation study of Orkla ASA notes that practitioners differ widely on whether to apply the discount and by how much, with estimates running from 0% to above 25%, depending on their view of how well management allocates capital. Treat any specific figure as an assumption and show its effect: run your valuation with 0%, 10% and 20% and see how much the conclusion moves.

Where SOTP goes wrong

Too many segments. More pieces give more room to assign a rich multiple to something small. Check that each split matches the filings.

Cherry-picked peers. Comparing a slow-growth division to the highest-multiple company in the sector is the easiest way to inflate value.

Ignored overhead and shared costs. Covered above, and the most common error in manual models.

Ignoring the cost of a break-up. SOTP shows theoretical value. Splitting a company costs money in taxes, advisers and lost shared services, and may not happen at all.

Treating the answer as a price target. Analysts do use SOTP to support optimistic targets, so read any sell-side SOTP as an argument, then check its inputs.

How to test the answer

The most accurate valuations rarely come from one method. A case study of investment bank analysts valuing a multidivisional German company found they used more than one method, mixing trading multiples with cash flow approaches. The analyst study cited earlier, which reviewed 265 equity reports, found that SOTP is often a preferred model but does not consistently beat a holistic DCF.

So cross-check:

  1. Run a consolidated DCF and compare it with your SOTP. If they are far apart, find out which assumption causes the gap.
  2. Check the implied blended multiple. If your SOTP implies the whole company deserves a multiple well above any peer, ask why.
  3. Compare with intrinsic value. Our intrinsic value calculation guide walks through a practical version.
  4. Look at other fundamentals. Profitability, balance sheet strength and cash generation still matter. A checklist of valuation metrics helps keep the view balanced.

SOTP and the Warren Buffett question

People often ask whether Buffett uses SOTP. The short answer is that it is not his main tool. As summarized in Lawrence Cunningham's essays collection, Buffett frames value as intrinsic value, the discounted cash that can be taken out of a business over its remaining life, and he prefers businesses whose earnings are predictable enough to estimate with confidence. His preferred yardstick is owner earnings: net income plus non-cash charges such as depreciation, minus the maintenance capital spending needed to stay competitive.

You can apply that thinking to each segment. Ask what cash each division produces after the spending needed to keep it competitive, and how durable it is. Our guide to calculating owner earnings shows how. That turns SOTP from a multiples exercise into a cash-based one.

Our perspective

We built Oracle Investments around a belief that investors should see how a number is made. Rule-based scoring on reported fundamentals, with the inputs visible, is the opposite of a black box. SOTP is a good example of why that matters.

A SOTP model has a dozen judgment calls: segment splits, multiples, overhead, discount. Each one can be defended, and together they can produce nearly any answer. That does not make the method bad. It means the answer is only as trustworthy as the inputs you can inspect.

Our perspective — overview diagram

Our advice for individual investors is practical. Use SOTP when the company truly is several businesses. Keep the segments to what is reported. Write down the multiple you used for each and why. Then check the result against something simpler: how profitable the company is, how it is valued compared with peers, and whether the balance sheet can take a shock. If a stock looks cheap on a SOTP but weak on basic fundamentals, the SOTP is usually the thing to doubt.

Value investing, in the tradition of Buffett, Munger and Lynch, rewards patience and simple checks more than elaborate spreadsheets. SOTP earns a place when it clarifies a complicated company, not when it adds complexity to justify a view you already hold.

— Matt

Using SOTP alongside a fundamentals screen

SOTP is slow. You read the segment note, pick peers, build the table and argue with yourself about multiples. That is worth doing for a handful of stocks you already like. It is not a way to scan a market.

For the scanning step, Oracle Investments scores over 260 stocks on profitability, valuation and financial health, so you can compare companies side by side and see which ones look potentially undervalued before you spend an evening on a segment model. The scores are rule-based and show their inputs, and the app also tracks your portfolio. Our stock comparison checklist shows how to line companies up the same way by hand.

A short checklist before you trust a SOTP

  • Do the segments match the company's reported segments?
  • Is each multiple drawn from peers that really resemble that segment?
  • Is corporate overhead charged to someone or capitalized as a negative?
  • Did net debt, pensions and other claims come out, and non-operating assets go in?
  • Have you tested the result at several discount levels?
  • Does a DCF or owner earnings estimate land in the same neighborhood?
  • Would you still like the stock if the company never breaks up?

The last question matters most. A SOTP gap closes only if something makes it close: a spin-off, a sale, an activist, or better capital allocation. Without a catalyst, a stock can stay below its parts for years. Experts often use SOTP as a pressure tool, for example to argue in a takeover fight that a company is worth more dismantled, forcing management to unlock value or defend the structure, as Investopedia notes. If you are a minority shareholder, ask who is going to do the pressing.

FAQ

What is the sum of the parts valuation method?

It values each business segment of a company separately and adds the results to get an enterprise value. You then subtract net debt and other claims and add non-operating assets to reach equity value. It is also called break-up analysis.

When should I use SOTP instead of a regular multiple?

Use it when a company owns businesses that deserve different multiples, such as a conglomerate or a holding company. For a focused company with one main business, a single multiple or a DCF is simpler and usually just as accurate.

Which multiple works best for SOTP?

EV/EBITDA is the most common, according to research on analyst practice. It is not always best: for capital-intensive segments, EV/EBIT or a segment DCF may reflect the economics better. Whatever you pick, use peers that resemble the segment.

Does Warren Buffett use sum of the parts?

Not as a primary tool. He focuses on intrinsic value and owner earnings, favoring businesses with predictable, durable cash flows. You can still borrow his approach by estimating owner earnings for each segment.

Is a SOTP value the same as a price target?

No. It is an estimate under a set of assumptions, and it ignores whether anything will release the gap between the parts and the stock price. Treat it as a range to test, not a forecast.

Sources