Back to Articles|Published on November 21, 2026|29 min read

Docket Article

Carve-Out Financial Due Diligence Checklist: Perimeter Gate

A 2026 buyer-side carve-out financial due diligence checklist: Rule 3-05(e) omitted costs, a blank perimeter matrix, a cost-inclusion bridge, TSA and dis-synergy rows, and a pass/pending/fail gate.

By Docket Research

Carve-Out Financial Due Diligence Checklist: Perimeter Gate

Summary

  1. 01Reconcile the reporting component, legal entities, proposed transfer of assets and liabilities, and future operating model against the mandate before using a public figure.
  2. 02A segment note uses the profit measure reported to the chief operating decision maker. An abbreviated statement includes depreciation and amortization and may omit corporate overhead, interest on debt excluded from the transfer, and income tax.
  3. 03For each cost class, document what the reported metric includes or omits and what independent operations would cost. Replace supported allocations once and track recurring costs, transition service charges, and one-time spending separately.
  4. 04Pass only when the perimeter reconciles, the metric matches the mandate, and the entire supported range clears the threshold. A wholly below-threshold range fails the numerical test, while an overlapping range or unbounded omitted costs leave the screen pending.
  5. 05Public revenue can support a provisional size screen once the perimeter reconciles, while profitability may remain pending until authorized diligence can establish the missing costs.
Inside this article
  1. 01Executive Summary
  2. 02Introduction and Background
  3. 03Key Changes: What the Rules and Standards Now Say
  4. 04The Four-Perimeter Crosswalk and Source-Type Taxonomy
  5. 05Reading the Notes: The Cost-Inclusion Bridge
  6. 06Applying the Acceptance Gate
  7. 07Data Analysis and Evidence
  8. 08Implications and Future Directions
  9. 09Frequently Asked Questions (FAQs)
  10. 10Conclusion

Executive Summary

A private-equity origination team screening a business that is being separated from a corporate parent usually meets one of five figures: a segment profit from a public note, a full carve-out statement, an abbreviated "revenues and direct expenses" statement, a non-GAAP deal presentation, or a management forecast. This report asks one question about each: does the public evidence support the mandate's revenue and profitability screen for the proposed perimeter, or must the answer stay pending until authorized diligence information exists? The short answer is that revenue can sometimes support a provisional size screen once the perimeter reconciles, while profitability very often cannot be accepted from public sources.

The main reason is structural. Current Rule 3-05(e) of SEC Regulation S-X lets a registrant present abbreviated statements for an acquired business that met all of four conditions, and those statements may omit corporate overhead, interest on debt not assumed and income tax ([1]). The notes must say the result is not indicative of the business going forward ([1]). The amendments were announced on May 21, 2020 and took effect January 1, 2021 ([2]), so they are not a 2026 development. The eCFR text was current through November 19, 2026 when retrieved ([1]).

A second reason is that history is not replacement cost. SEC staff guidance expects historical statements to reflect all costs of doing business ([3]), but allocated amounts "may not be reflective of the actual costs expected to be incurred" ([4]). Segment profit is a third trap: it is the measure given to the chief operating decision maker, including allocations only if that measure includes them ([5]).

Published practitioner ranges show why the gap matters, though they are context and never inputs: BCG reports one-time separation costs of about 1% to 5% of divested revenue and up to 13% in large, complex cases ([6]), and Roland Berger's April 2026 survey reports average ongoing cost increases of 11 percent after independence ([7]). No public source supplies a cost for any particular business. The recommended control is a source-to-metric acceptance gate built on two blank worksheets, a four-perimeter matrix and a cost-inclusion bridge. A numerical pass requires the whole supported range to clear the threshold, a fail requires the whole range to fall below it, and everything else is pending, and a separate perimeter gate can block a numerical pass. This is origination research, not accounting or legal advice, and consequential interpretations belong with qualified specialists.

20 percentCeiling for the acquired business's assets and revenues as a share of the seller's consolidated amounts under the abbreviated statement rule
1% to 5%BCG's approximate range for one-time separation costs as a share of divested revenue
13%One-time separation costs as a share of divested revenue in some large, complex cases in BCG's analysis
11 percentAverage increase in ongoing costs after independence reported in Roland Berger's survey

Introduction and Background

Origination analysts rarely receive a clean answer to a simple question: what are the annual external revenue and the defined profit of the business component a parent might separate? The mandate states a metric (for example, annual external revenue above a floor and operating profit above a floor), a period, a currency, a set of business activities and a geography. Public documents describe something close to that, but rarely the same thing. A segment, a product line and the eventual sale perimeter can all differ.

This report is written for the buyer's screening decision. It does not identify an acquisition opportunity, prepare financial statements, value a transaction or assert that securities-filing rules govern every private deal. Rule 3-05 is a disclosure framework for registrants that acquire businesses: its text addresses filings of acquired-business statements and refers to debt that "will not be assumed by the registrant" ([1]). A private seller and a private buyer are not bound by it, although the documents it produces are a useful public analogue when they exist. Readers who first need the broader source-quality gate should start with the revenue-estimation guide and the Exhibit 21 corporate-family guide, which are prerequisites here: this report assumes a revenue figure has already been traced to its source and focuses on what that source covers.

Existing professional guidance is deep but points elsewhere. EY's accounting guide to carve-out financial statements was updated on April 29, 2026 ([8]). EY-Parthenon's September 2026 article addresses how a seller builds a standalone cost analysis and operating model ([9]). KPMG's IFRS discussion, whose body is dated May 2017 despite a 2023 URL path, notes that no IFRS literature specifically defines combined or carve-out statements ([10]). Those are preparation guides. The gap this report addresses is the buyer's acceptance test: given a public number, which perimeter and which costs does it describe, and what remains unknown?

Three principles organize the rest of the analysis:

  • Public absence is unknown, not zero. If a note does not state an allocation basis, the allocation is unknown rather than absent.
  • Four perimeters stay separate. The financial reporting component, the legal-entity family, the proposed transferred assets and liabilities, and the modeled future operating model are different objects.
  • Each cost is counted once. A historical allocation, an omitted expense and a replacement cost occupy different rows of one bridge.

The analysis is dated as of November 2026.

Key Changes: What the Rules and Standards Now Say

The regulatory and standard-setting text that shapes public carve-out evidence changed in two main places between 2020 and 2025: the SEC's 2020 acquired-business amendments and FASB's ASU 2023-07 on segment expenses. Each change alters what a buyer can infer from a public figure.

The 2020 amendments to acquired-business reporting

The SEC announced its amendments to financial disclosures about acquired and disposed businesses on May 21, 2020, with an effective date of January 1, 2021 and voluntary earlier compliance permitted ([2]) ([2]). The fact sheet says acquired-business statements now cover no more than the two most recent fiscal years ([2]). The same release introduced the permission to omit certain expenses for acquisitions of a component of an entity ([2]).

Two cautions follow. First, an article describing this framework in 2026 should not call it new. Second, the adopting release states that "neither our proposal nor the final rule addresses carve-out financial statements" ([11]), leaving full carve-out statements to staff consultation on the facts. Abbreviated statements and carve-out statements are therefore different instruments, and a buyer should record which one it holds.

Significance tests set how much history a registrant must supply. If no significance condition exceeds 20 percent, separate statements are not required ([1]). Above 40 percent, the rule calls for at least the two most recent fiscal years ([1]). Rule 3-05 applies the significant-subsidiary tests through Rule 11-01(b), with 20 percent substituted for 10 percent each place it appears ([12]). Under the investment test, the registrant's investments in and advances to the acquired business are compared with the aggregate worldwide market value of its common equity, or with total assets if there is no such market value ([13]). The base definition uses 10 percent; the 20 percent figure is the one that governs the tiers above. For a buyer, the practical point is narrow: whether any statement exists at all depends on the registrant's significance, not on the target's attractiveness to the mandate.

Abbreviated statements under Rule 3-05(e)

Paragraph (e) allows abbreviated presentation for an acquisition of net assets that constitutes a business, such as a product line, only if all qualifying conditions are met. The total assets and revenues, after intercompany eliminations, must be 20 percent or less of the seller's consolidated amounts ([1]). Critically, the business must not have been "a separate entity, subsidiary, operating segment" or division during the periods required ([1]).

That last condition has a direct screening consequence. On its face, a business that already appears as a reported segment in a public note is not the kind of business paragraph (e) addresses. A buyer holding a segment note and an abbreviated statement for what looks like the same activity should treat the pair as a perimeter question to resolve, not as two confirmations of one number.

The permitted omissions are limited. The statement of comprehensive income must include expenses incurred by or on behalf of the business, including cost of sales, selling, general and administrative expenses, depreciation and amortization, and research and development, but may omit corporate overhead, interest on debt not assumed and income tax expense ([1]). The title must signal the omission ([1]), and the notes must describe the omitted costs, explain why full statements are impracticable and explain that the result is not indicative going forward ([1]). The notes must also give operating, investing and financing cash flows to the extent available ([1]).

Two arithmetic implications matter. An abbreviated result already deducts depreciation and amortization, so an analyst computing EBITDA must find that expense in the notes rather than assume it. And because corporate overhead may be omitted, the abbreviated figure can overstate profit relative to an independently run business. The SEC staff's Financial Reporting Manual describes the same boundary, permitting exclusion of costs "such as corporate overhead, interest and taxes" ([14]). That manual page carries a 2017 edition date, so it is useful as a description of staff thinking but should be checked against the live manual before it is relied upon, and its older discussion of consulting staff before filing is not blanket advice for private buyers.

Pro forma presentation and autonomous-entity adjustments

Article 11 pro forma information is a third kind of number. It applies an "autonomous entity" presentation when the registrant previously was part of another entity ([12]). Autonomous Entity Adjustments appear in a separate column from Transaction Accounting Adjustments ([15]), and management may optionally add synergy and dis-synergy adjustments. When synergies are presented, related dis-synergies must also be presented ([15]), and an expense reduction must not exceed the related expense historically incurred ([15]).

For screening, a pro forma figure belongs in its own row of the source-type taxonomy. It describes a registrant after a defined transaction, under stated assumptions, and is not an audited historical carve-out result. The SEC staff manual also warns that statements for carved-out businesses necessarily exclude various operating expenses ([16]).

Segment disclosure after ASU 2023-07

FASB's Accounting Standards Update 2023-07 added the significant expense principle: public entities must disclose significant segment expenses regularly provided to the chief operating decision maker ([17]). It applies to fiscal years beginning after December 15, 2023 and interim periods in fiscal years beginning after December 15, 2024 ([17]), so most public segment notes for 2025 and 2026 should now show more expense detail than older filings.

The update also requires disclosure of an "other segment items" amount ([17]) and of the nature and effect of any asymmetrical allocations to segments ([17]). Allocated amounts in segment measures must be on a reasonable basis ([17]). None of this converts segment profit into a standalone profit. The IFRS 8 equivalent likewise defines the measure as the one reported to the chief operating decision maker ([5]).

Figure 01
Acquired-business reporting and segment disclosure dates
  1. May 21, 2020SEC amendments announced

    The SEC announced amendments to financial disclosures about acquired and disposed businesses.

  2. Jan 1, 2021SEC amendments take effect

    The acquired-business reporting amendments took effect. The SEC permitted voluntary early compliance.

  3. Dec 15, 2023Fiscal-year application

    ASU 2023-07 applies to fiscal years beginning after this date.

  4. Dec 15, 2024Interim-period application

    ASU 2023-07 applies to interim periods in fiscal years beginning after this date.

A segment, a product line and the eventual sale perimeter can all differ.

The Four-Perimeter Crosswalk and Source-Type Taxonomy

Before reading any number, record what kind of document it comes from. The SEC's non-GAAP interpretations draw a useful contrast: segment profit that GAAP requires is not itself a non-GAAP measure, being "the measure reported to the chief operating decision maker" ([18]), while a deal-basis "Adjusted EBITDA" is a different animal that must be reconciled and labeled differently.

Source types to separate

  • Segment note: CODM-basis measure, may include or exclude allocated corporate costs, may include intersegment revenue.
  • Full carve-out or combined statements: reporting perimeter set by purpose rather than legal form, with allocated common costs.
  • Rule 3-05(e) abbreviated statements: revenue and direct expenses with omitted overhead, interest and tax.
  • Non-GAAP deal-basis presentation: management-defined measure that needs its own reconciliation.
  • Pro forma presentation: registrant-level result under stated transaction and autonomous-entity assumptions.
  • Management forecast: prospective, unaudited, outside the historical evidence set.

For each source type, retain the accounting basis, the assurance scope, the cutoff date, the document key and the limits stated in the notes.

Why the reporting perimeter is not the legal-entity family

KPMG's IFRS discussion states that the entity in combined statements is "tailored to meet the specific purpose for which the financial statements are prepared" ([10]). Combined or carve-out information does not represent an existing legal entity or group ([19]). Preparers may build it top-down from the larger entity's consolidated statements ([10]) or bottom-up from legal-entity ledgers.

The same logic defeats the shortcut of using a subsidiary list. Exhibit 21 requires a list of subsidiaries and their jurisdictions ([20]), and names may be omitted when the unnamed subsidiaries together would not be a significant subsidiary ([20]). It is a legal-entity list. It says nothing about which activities, assets or liabilities form the financial business being separated.

Table 1: blank perimeter matrix

Table 1 below is the first reference asset: a matrix with one row per source and every cell blank until a document supports it. The rule for filling it is simple: a cell holds either a sourced fact with its note reference or the word Unknown.

FieldWhat to recordEvidence required before fillingIf absent
Source and document keyAccession number, page or note referenceRetrieved documentNot usable
Source typeSegment note, carve-out, abbreviated, deal-basis, pro forma or forecastDocument title and basis statementUnknown
Assurance and basisAudit or review scope, accounting frameworkReport and basis-of-presentation noteUnknown
Periods and cutoffFiscal years, interim periods, date of last updateStatement headersUnknown
Revenue definitionExternal versus intersegment, gross or netRevenue and related-party notesUnknown
Reporting componentSegment, product line, division, combined entitiesBasis-of-presentation noteUnknown
Legal entitiesEntities included or excludedEntity list tied to the component, not an Exhibit 21 list aloneUnknown
Included activitiesProducts, geographies, customersSegment or business description noteUnknown
Transferred or retained assets and liabilitiesShared assets, debt, contingenciesBalance-sheet notesUnknown
Intercompany treatmentEliminated, retained as related-party, or ceasedRelated-party noteUnknown
Difference from mandateActivity, period, currency or geography gapsSide-by-side comparisonUnresolved

The matrix is deliberately plain. Its value is the Unknown column: the most common failure in screening is an unrecorded assumption that a revenue figure covers the mandate's perimeter. PwC's carve-out guidance supplies the reasoning for the balance-sheet rows: the carve-out balance sheet includes assets that relate to the operations of the business ([21]), shared long-lived assets are frequently shared resources that must be judged as a whole ([21]), and parent debt is generally not reflected ([22]). Intercompany transactions between entities inside the carve-out remain eliminated ([23]), while transactions with the parent that were eliminated at group level become related-party items requiring separate disclosure ([23]).

Figure 02
Segment notes and abbreviated statements
Segment note
  • Uses the profit measure reported to the chief operating decision maker, with allocations only if that measure includes them.
  • Reported revenue may include intersegment sales.
Abbreviated statement
  • The reported result already deducts depreciation and amortization.
  • The statement may omit corporate overhead, interest on debt the buyer does not assume, and income tax expense.

If both documents appear to cover the same activity, resolve the perimeter before treating them as corroborating evidence.

Reading the Notes: The Cost-Inclusion Bridge

The matrix fixes the perimeter. The bridge fixes the costs. Its purpose is to prevent three arithmetic errors: subtracting an included allocation twice, adding back a cost without evidence that it disappears, and mixing recurring, time-limited and one-time items.

What historical statements include

SEC staff guidance states the baseline principle that historical statements should reflect all costs of doing business ([3]). Where specific identification is not practicable, a reasonable allocation method is expected ([3]), with a note explaining the method. PwC's guide lists typical bases, such as time, headcount, square footage or usage, and states that the guidance is not prescriptive ([4]). KPMG notes that allocation by size measures such as revenue or assets is likely to be inappropriate unless size correlates directly with the expense ([19]).

Allocated parent costs are related-party transactions, so related-party disclosures appear in the notes ([4]). Those notes are where a buyer finds the allocation basis, if it exists.

Why an allocation is not a replacement cost

The SEC staff has described a stand-alone estimate as the cost that would have been incurred had the subsidiary operated as an unaffiliated entity ([3]), and PwC adds that the allocated costs may not reflect costs expected after separation ([4]). Deloitte's guidance for divestiture planning likewise asks both sides to understand how shared-service costs were historically charged and how they would differ on a stand-alone basis ([24]). EY-Parthenon says a parent-era cost structure rarely fits a newly independent company ([9]).

Dis-synergies explain part of the difference. BCG describes them as higher unit costs as economies of scale decrease ([6]). Deloitte's Swiss guidance observes that where allocation keys fail to recharge IT costs fully to the divested entity, the buyer can inherit an understated cost base ([25]).

Table 2: blank cost-inclusion bridge

Table 2 below is the second reference asset. One row per cost class (for example executive management, finance, legal, information technology, human resources, procurement, facilities, insurance, brand and royalties, research support). Cells stay blank until a source note fills them.

ColumnQuestionAllowed entriesErrors it prevents
Cost classWhich shared or parent-provided cost?Named categoryOmitting a class entirely
Included in reported metric?Is the cost inside the figure being screened?Yes, No or UnknownSubtracting twice
Allocation basisHow was it charged?Basis named in a note, or UnknownTreating all allocations as comparable
Omitted reasonWhy is it excluded (for example corporate overhead under the abbreviated rule)?Note referenceAssuming exclusion means the cost vanishes
Replacement amount or rangeProspective cost to run independentlySourced figure, range or UnknownImporting a generic benchmark
NatureRecurring, time-limited TSA or one-timeOne valueMixing cost types
Expiry or transition assumptionWhen does a TSA charge stop?Contract or note reference, or hypothesisAssuming a TSA end date
Source and dateWhere does the entry come from?Document keyUntraceable adjustments
ConfidenceSupported, bounded or hypothesisOne labelPresenting hypotheses as facts
Effect on mandate metricRaises, lowers, or unknownDirection and amount if supportedPremature pass

The bridge works because it separates three questions that deal presentations tend to fuse. First, what did the reported metric already charge? Second, what did it omit by rule or by choice? Third, what will a stand-alone operation plausibly cost, and for how long? Only the first can usually be read from a public filing, and only if the allocation note is specific.

Distinguishing recurring, time-limited and one-time costs

Transition services agreements (TSAs) sit between the historical charge and the stand-alone cost. BCG reports that TSAs typically run 3 to 24 months after closing ([6]), and that the buyer usually pays for ongoing TSA services ([6]). PwC states a typical TSA exit takes 12 to 24 months, with an expedited approach reaching 6 to 12 months ([26]). Deloitte's guidance for financial buyers notes that sellers often apply a modest single-digit markup over the actual cost of services ([27]), and that one-time TSA items such as data migration needed for exit belong in the buyer's one-time costs ([27]).

These are practitioner observations, not contract terms for any target. A buyer must keep three buckets apart: recurring replacement cost (continues indefinitely), time-limited TSA charges (stop at an exit date that may slip), and one-time separation expenditure (BCG describes these as the parent's costs to separate operations and assets ([6])). Deloitte's South African guidance recommends an agreed exit process for each TSA prepared early ([28]), a reminder that an expiry date is a hypothesis until contracted.

Arithmetic rules for the bridge

  • Replace an included allocation once. If the reported metric already carries an allocation, add it back and then deduct the supported prospective cost; do not simply deduct the prospective cost on top.
  • Do not add back without disappearance evidence. A cost is a candidate add-back only if a note or contract shows it ends. The SEC's non-GAAP interpretations warn that excluding "normal, recurring, cash operating expenses" can make a measure misleading ([18]), and that expenses occurring repeatedly or occasionally are viewed as recurring ([18]).
  • Never fill an omitted class with a ratio. Allocating parent revenue by employee count to fill a perimeter gap converts an unknown into a false precision.
  • Keep absence as unknown. A blank is not zero.
  • Cap pro forma synergies at the historical expense. That is the rule's own limit for management adjustments ([15]).
Figure 03
Build the cost-inclusion bridge
  1. 01Identify included costs

    Record whether each cost class is included and document the allocation basis from the note. Mark unsupported entries Unknown.

  2. 02Document omitted costs

    Record what the source omitted by rule or choice and capture the reason from the note.

  3. 03Check the add-back

    Use a note or contract to verify that an existing cost ends before adding it back.

  4. 04Replace allocations once

    When adjusting an included allocation, add it back once and deduct the supported prospective cost.

  5. 05Classify costs by duration

    Track recurring replacement costs, time-limited transition services charges, and one-time separation spending separately.

Applying the Acceptance Gate

The gate combines a numerical test with a perimeter test. The SEC staff's manual describes a carve-out as acceptable where a reasonable basis exists to allocate what cannot be specifically identified ([14]), which is the same logic: an allocation is a judgment with a stated basis, not an observed fact. The numerical test asks whether the whole supported range of the mandate metric clears the threshold. The perimeter test asks whether the metric describes the mandate's perimeter at all.

Five outcomes

  1. Scope matched and supported metric. The perimeter matrix reconciles, the metric definition matches the mandate, and the supported range clears the threshold.
  2. Supported with bounded assumptions. The metric is supported, but a documented, bounded assumption (for example a contracted TSA charge) is needed.
  3. Threshold straddles the range. The supported range includes values both above and below the threshold. The answer is pending.
  4. Omitted costs unbounded. A cost class is omitted by rule or choice and no source bounds its replacement cost. The answer is pending.
  5. Perimeter unresolved. The reporting component, entities or assets cannot be tied to the mandate. No numerical pass is available regardless of the figure.

A numerical pass requires the entire supported range to clear the threshold. A fail requires the entire range to be below it. Anything else is pending. The consequence is a deliberately uncomfortable asymmetry: revenue disclosed in a segment note may support a provisional size screen under outcome 1 or 2 while profitability remains pending under outcome 4, because revenue is less exposed to omitted overhead than profit.

Defining the profitability field first

Verify the definition of any profit measure before computing with it. In SEC-registrant filings, the staff's non-GAAP interpretations state that "earnings" in EBIT and EBITDA means GAAP net income as presented in the statement of operations ([18]), that operating income is not the most directly comparable measure ([18]), and that measures calculated differently should not carry the EBIT or EBITDA label ([18]). These interpretations govern registrant filings, not private deal documents, so they work here as a labelling and reconciliation test for the analyst. A carve-out "EBITDA" in a deal document may be an adjusted measure with its own add-backs, and the analyst must locate its reconciliation.

Three metrics also need to stay apart. Accounting profit is a historical measure under a stated basis. Pro forma presentation applies stated transaction and standalone assumptions to a registrant. Ongoing cash requirement includes TSA fees, separation spending and working-capital needs that no profit measure shows. A business can pass an accounting-profit screen and still absorb a TSA-period cash demand, which is why the bridge labels each row recurring, time-limited or one-time.

Table 3: fictional sensitivity grid

Table 3 below is purely fictional and carries no benchmark. It shows only how transition duration changes the cumulative shared-service cost when a replacement cost differs from the historical charge. (Hypothetical Example) Assume a fictional carve-out component whose historical shared-service allocation is 9 units a year, a hypothetical TSA charged at that historical rate, and a prospective replacement cost applied after the TSA ends. The grid shows 24-month cumulative shared-service cost, in fictional units, for each replacement cost and TSA duration. The inputs are hypotheses with no source.

Annual replacement cost (units)TSA 6 monthsTSA 12 monthsTSA 18 monthsTSA 24 months
816.517.017.518.0
1222.521.019.518.0
1628.525.021.518.0
2034.529.023.518.0
2440.533.025.518.0

The grid isolates a point that is easy to miss: when replacement cost is above the historical charge, a longer TSA can lower the 24-month cumulative figure while only deferring the higher run-rate cost; when replacement cost is below the historical charge, the opposite occurs. It also excludes one-time separation spending entirely. A reviewer should read it as a model of arithmetic, not a forecast, and should not copy any cell into a deal record.

An illustrative bridge (Hypothetical Example)

Consider a fictional segment, "Alpha Components Unit", with a mandate threshold of 50 units of operating profit. Its fictional segment note reports 62 units after an allocated shared-service charge of 9 units, which the note identifies as included. Suppose a fictional contract and cost note support a prospective replacement cost of 12 to 17 units a year.

  1. Start with the reported metric: 62.
  2. Add back the included allocation once: 62 plus 9 equals 71.
  3. Deduct the supported replacement range: 71 less 17 equals 54, and 71 less 12 equals 59.
  4. Compare: the whole supported range, 54 to 59, clears the threshold of 50.

The result is a model, not filed EBITDA, and it does not reach outcome 1 unless the perimeter gate also passes. If the same fictional note had instead omitted corporate overhead with no replacement estimate, outcome 4 would apply and the answer would be pending, even though the reported 62 looks comfortable. If the supported range had been 45 to 55, the threshold would straddle it, producing outcome 3. In every variant the inputs are invented to show the logic; the example makes no claim about any real business.

Figure 04
Apply the source-to-metric acceptance gate
  1. 01Reconcile the perimeter

    Tie the reporting component, entities, and assets to the mandate before accepting a numerical pass.

  2. 02Check the metric definition

    Confirm that the source uses the mandate's metric definition. Locate the reconciliation for any adjusted deal-basis EBITDA.

  3. 03Bound the supported range

    Document any bounded assumptions. Keep profitability pending when an omitted cost has no sourced replacement range or amount.

  4. 04Compare the full range

    Compare the entire supported range with the threshold. A range that straddles the threshold leaves the screen pending.

The perimeter reconciles, the metric matches the mandate, and the entire supported range clears the threshold.

The entire supported range falls below the threshold.

Data Analysis and Evidence

Quantitative evidence in this field is mostly practitioner survey and consulting analysis, rather than regulator or peer-reviewed statistics, and none of it is a benchmark for any particular business. The figures below are attributed to their originators and are offered to scale the unknowns, not to fill them.

Regulatory thresholds and dates

  • 20 percent, 40 percent: significance tiers in Rule 3-05(b) that govern whether one or two fiscal years of acquired-business statements are required ([1]).
  • 20 percent of assets and revenues: ceiling for abbreviated statements under paragraph (e) ([1]). In its adopting release the SEC explained that a 20 percent level identifies when an acquired business is a small portion of the seller ([11]).
  • December 15, 2023 and December 15, 2024: the fiscal-year and interim-period effective points for ASU 2023-07 segment expense disclosure ([17]).

Published cost and duration ranges

BCG compiled separation cost data on more than 50 divestitures ([6]), reporting one-time costs of roughly 1% to 5% of divested revenue and up to 13% in some large, complex cases ([6]). Roland Berger's April 2026 survey of more than 50 senior carve-out practitioners reports average ongoing-cost increases of 11 percent after independence ([7]) and one-time separation costs of up to roughly 15 percent of revenue in highly entangled cases (the release words this both as "exceed 15 percent" and "up to 15 percent") ([7]). The two sources use different samples, definitions and years, so their one-time ranges should not be combined into a single band. Roland Berger's figures come from a press release, not the full study.

On duration, BCG states 3 to 24 months for TSAs ([6]), while PwC states 12 to 24 months as a typical exit and 6 to 12 months for an expedited approach ([26]). PwC adds that carve-out buyers often overrun their one-time cost and EBITDA budgets ([26]). These are practitioner observations and sit alongside, not inside, the fictional grid above.

Market and return context

Bain, using DealEdge data, reports that the average private-equity carve-out since 2012 earned a 1.5x multiple of invested capital ([29]) and that few carve-outs have standalone financial reporting, which makes a clear picture of profitability difficult ([29]). Bain's best-practice analysis covered 25 carve-outs completed between 2013 and 2024 ([29]). McKinsey puts buy-side carve-outs at 28 percent of M&A transactions above a size threshold over a stated period ([30]) and cautions that a seller may prepare standalone financials to be as marketable as possible ([30]). Academic evidence on seller outcomes is mixed: one working paper reports that private equity accounted for about 30% of large corporate asset sales in its sample ([31]), while a 2024 journal article associates divestitures to private-equity acquirers with lower divesting-firm shareholder returns on average ([32]).

Count each cost once and label it recurring, time-limited or one-time.

Implications and Future Directions

A next evidence request for the authorized deal team

If a source-to-metric review leaves the answer pending, the follow-up belongs with the authorized deal team under whatever confidentiality terms apply, rather than with the seller by way of unsolicited outreach. A compact request, framed as questions rather than conclusions, covers four areas:

  • Scope and metric: What period, currency and revenue definition does the figure use, and does it include intersegment or related-party revenue?
  • Included and omitted costs: Which cost classes were allocated, on what basis, and which were omitted by rule or choice?
  • Allocation basis: Is the allocation methodology documented in a note, and has it changed between periods?
  • Standalone bridge: Does any document estimate the cost of running the business independently, with a stated operating model, and what TSA terms and expiry assumptions exist?

Financial and legal specialists, not the origination analyst, govern how those answers are interpreted.

Valuation pitfalls to avoid at the screening stage

This report does not value any transaction, but several valuation errors begin in the screening record. Do not apply a generic multiple to an abbreviated profit figure whose omitted costs are unbounded, since the figure is described by its own rule as not indicative going forward. Do not infer private-company margins from public segment margins. Do not treat a seller-prepared standalone estimate as neutral: McKinsey notes that such estimates can be prepared to be as marketable as possible ([30]), and EY notes that where no seller estimate exists buyers will likely fill gaps with their own assumptions ([9]). Bain's screening question for sponsors is whether they truly know how much profit the company could generate on its own ([29]). Until the bridge can answer it, the record should say so.

What to watch

Three developments deserve monitoring:

  • Segment expense detail: segment notes in 2026 filings will show more expense detail than earlier years under ASU 2023-07, which may bound some allocation questions earlier than before.
  • Professional guidance updates: EY's guide update of April 2026 is one example.
  • Dated staff material: the SEC staff manual pages consulted for this report carry 2017 dates, so any reliance on them should be rechecked against the live manual.

Because KPMG's IFRS discussion shows that cross-border reporting judgments differ from a US presentation, an analyst should also not assume that a US rule recipe settles a non-US perimeter question.

Frequently Asked Questions (FAQs)

What does a carve-out due diligence worksheet need at minimum?

Two blank matrices: the perimeter matrix (Table 1) and the cost-inclusion bridge (Table 2), with every cell filled from a source note or marked Unknown. The gate outcome then follows from the matrices.

Can an abbreviated statement be used as EBITDA?

Not without further work. It includes depreciation and amortization and other operating costs but may omit corporate overhead ([1]), and, in SEC-registrant filings, the staff's non-GAAP interpretations treat EBITDA as starting from GAAP net income (a labelling and reconciliation test for the analyst, not a rule on private deal documents) ([18]).

What are dis-synergies in a carve-out?

Higher unit costs when scale is lost ([6]). Pro forma presentation requires related dis-synergies to be shown if synergies are shown ([15]).

How long do TSAs last and who pays?

Practitioner sources describe 3 to 24 months, with the buyer usually paying for ongoing services ([6]). Contract terms for a specific deal are not public.

Which carve-out deal screening criteria are worth adding?

Beyond revenue and profit thresholds, add the perimeter test, the definition test for any EBITDA figure, and a flag for unbounded omitted costs. Bain notes few carve-outs have standalone financial reporting ([29]), so many screens will legitimately end as pending.

Do these securities rules apply to every private deal?

No. Rule 3-05 is a disclosure framework for registrants and filings, and this report uses it as a public analogue only. Whether any rule applies to a specific transaction is a matter for legal counsel.

Conclusion

The central finding is that a public figure should be accepted into an acquisition record only through a documented path from source to metric. A segment profit, an abbreviated result, a pro forma column and a management forecast each answer a different question, and none of them is a complete standalone cost base. Current Rule 3-05(e) permits specified omissions and requires notes warning that the presentation is not indicative of future operations. SEC staff guidance and professional literature treat historical allocation and stand-alone cost as different concepts. Segment standards define profit by what is reported to the chief operating decision maker.

For screening, this yields a clear discipline. Fill the perimeter matrix before reading numbers. Fill the cost-inclusion bridge before computing profit. Count each cost once and label it recurring, time-limited or one-time. Pass only when the whole supported range clears the threshold, fail only when it is wholly below, and otherwise hold the answer as pending. Revenue may support a provisional size screen while profitability waits for authorized management information, and a perimeter that cannot be reconciled blocks a numerical pass in either case.

That outcome is not a failure of research. It is a usable record of what public evidence can and cannot support, which is more valuable to a deal team than an unsupported number that later has to be unwound. The analysis here is for origination research; consequential accounting or legal interpretations require review by qualified professionals.

External Sources (32)

Published by

Docket Research

Docket provides deal-origination research software and managed research for private equity teams, screening companies against a mandate and showing the evidence behind each finding.

Our platform · Evidence standards · Contact

Get your first 10 targets free

Editorial note

Researched and written by Docket Research with AI-assisted research tools; sources are cited inline and checked before publication. Report an error: https://docket.capital/evidence-standards#corrections.