Document type: Guide Practice area: Corporate — Mergers and Acquisitions Jurisdiction: United States (federal and state), with cross-border notes Last reviewed: 5 September 2026

Phase 0: Seller readiness, before a buyer exists

The single highest-return investment in a carve-out is work the seller does before the process starts.

Define the perimeter internally. What is the business? Which product lines, which customers, which plants, which people, which contracts, which intellectual property. Write it down and circulate it, because the version in each executive's head differs.

Produce carve-out financial statements early, and have them reviewed or audited if the process warrants. Buyers discount uncertainty, and financial statements produced under deal pressure produce more of it.

Build the seller's own standalone cost estimate, so that the seller is not hearing the number for the first time from the buyer's adviser.

Model stranded costs. How much of the shared services organization does the business absorb, and how much can actually be eliminated? A divestiture that improves reported margin and degrades actual margin is a familiar and avoidable outcome.

Inventory shared contracts and inbound licenses. Categorize by treatment path and flag every agreement with a transfer, assignment, or change-of-control restriction. This inventory determines the structure, and building it late means choosing a structure that breaches things.

Draft an IT separation approach. Clone-and-divide, extract-and-implement, or continue-on-seller-systems. The choice sets the TSA's length, which sets a material part of its cost.

Stand up a separation management office with a named leader, functional workstreams, an integrated plan, and a weekly steering committee that can decide things.

Do the retention planning. Announcement day is when the recruiters call.

Phase 1: Choose the structure

Three options, evaluated against four criteria — tax cost, consent burden, timetable, and post-closing cleanliness.

Equity sale of existing entities. Available only where the business already sits in dedicated entities. Lowest consent burden for contracts (which stay with the entity), though change-of-control clauses may still bite. Rarely available.

Asset sale. Every asset and liability transfers individually. Highest consent burden, longest schedules, and the most opportunity for something to be left behind — but no pre-closing reorganization and no reorganization tax.

Reorganization-then-equity-sale. The seller forms new entities, transfers the business in, then sells the equity. The default for large multi-jurisdiction carve-outs. It converts a complex asset transfer into a simple equity sale.

But price the reorganization before committing to it.

Tax. Every internal transfer has consequences in every jurisdiction: transfer taxes, VAT, stamp duty, capital gains on intercompany transfers, withholding. The reorganization's tax cost can exceed its negotiating value.

Licenses. This is the trap. Federal common law treats patent licenses as personal and non-assignable absent consent — PPG Industries, Inc. v. Guardian Industries Corp., 597 F.2d 1090 (6th Cir. 1979) — and Cincom Systems, Inc. v. Novelis Corp., 581 F.3d 431 (6th Cir. 2009) held that an internal reorganization breached a software license, with no external assignment at all. Trademark licenses follow similar reasoning, as in In re XMH Corp., 647 F.3d 690 (7th Cir. 2011). Review every material inbound license for transfer language before choosing the structure.

Employment. In automatic-transfer jurisdictions, each internal transfer may itself trigger transfer and consultation obligations — sometimes twice, once into the new entity and once on the sale.

Regulatory. Permits and registrations held by the transferring entity may need reissuance at the reorganization step and again at closing.

Consider a spin-off where the goal is a distribution to shareholders rather than a sale: 26 U.S.C. § 355 can provide tax-free treatment if the active trade or business, device, business purpose, and continuity requirements are satisfied — with anti-abuse rules that punish a spin-off followed by a prearranged acquisition.

Phase 2: Triage the contracts

Every shared contract goes into one of four buckets. Do this on a schedule, with a named owner and a target date for each.

Bucket 1 — Assign in whole. Predominantly the divested business's. Requires counterparty consent in most cases; add a license-back or services arrangement if the seller still needs anything.

Bucket 2 — Split or partially assign. Two contracts from one. Requires the counterparty's active cooperation, which it will price. Use where the relationship matters to both sides.

Bucket 3 — Retain and pass through. The seller continues to perform and passes the benefit to the buyer. Check that the arrangement does not itself breach the contract, and set a sunset.

Bucket 4 — Terminate and replace. The buyer contracts on its own. Cleanest, and usually more expensive because the buyer buys at its own scale. Quantify the pricing difference and put it in the model — it is a real dis-synergy.

Run consent diligence before signing. Categorize by whether consent is required, whether the clause captures a change of control as well as an assignment, and how material the contract is. Counterparties who learn of the transaction and the timetable simultaneously will price accordingly.

In the purchase agreement, provide for: allocation of consent costs; alternative arrangements where consent is not obtained (benefit held for the buyer, performance at the buyer's direction and cost); a sunset on those arrangements; and a consequence only for contracts specifically identified as material.

Phase 3: Intellectual property

Get the formalities right. A written assignment recorded under 35 U.S.C. § 261. Trademark assignments including the goodwill as 15 U.S.C. § 1060 requires — an assignment in gross is invalid. Copyright transfers in a signed writing under 17 U.S.C. § 204.

Build three schedules.

Transferred IP. Registrations by number and jurisdiction; unregistered marks; key trade secrets and know-how; software; domain names; and social media accounts.

Retained IP licensed to the buyer. Field-of-use, exclusivity, sublicensing, improvements, term, and territory.

Transferred IP licensed back to the seller. The mirror image, and equally important.

Handle the brand. A transitional trademark license permitting the divested business to keep using the parent's name on products, packaging, signage, domains, and in foreign entity names — with quality control provisions, a rebranding plan with milestones, and a hard sunset (twelve to twenty-four months is typical).

Inbound licenses. For each, decide: does it transfer; is consent needed; will the licensor reprice; is there a replacement. The ones the business cannot operate without are leverage in the licensor's hands, and the price should be negotiated early rather than under closing pressure.

Open source. The divested business's code may include components with obligations that were satisfied at the parent level. Re-run the analysis for the standalone entity.

Phase 4: The people workstream

Identify the population. Dedicated employees first; then the hard cases — shared services personnel, matrixed managers, and people who split time. Expect a multi-week negotiation over a list of names, and expect the buyer to want more of the shared population than the seller wants to give.

In offer-and-rehire jurisdictions (including the United States).

  • Buyer makes offers on agreed terms: comparable base and bonus opportunity, prior service credit for vesting and eligibility, and severance protection for a period.
  • Analyze WARN, 29 U.S.C. § 2101, and state mini-WARN statutes, before the list is final — whether a group counts as terminated can depend on how many receive offers.
  • Allocate health continuation responsibility under 29 U.S.C. § 1161 for employees not hired.
  • Plan qualified plan treatment: a 401(k) spin-off, a transfer of assets and liabilities, or a distributable event — with the controlled group and plan definitions of 26 U.S.C. § 414 governing the analysis.
  • Address non-qualified deferred compensation, outstanding equity awards, accrued vacation, and retention arrangements individually.

In automatic-transfer jurisdictions. Employment transfers by operation of law with terms and accrued rights intact; information and consultation obligations attach; and the consultation timetable is frequently the transaction's critical path. Start it as early as the process permits.

Retention. Fund the pool before announcement. Allocate cost between seller and buyer. Include clawback for voluntary departure within a period.

Leadership. The divested business often has no CEO, CFO, or general counsel of its own because the parent provided them. Someone has to hire them, and for a financial buyer that search should begin before signing.

Phase 5: Build the standalone cost model

This is the diligence exercise that determines price, and it is the one buyers most often shortchange.

Start from the carve-out statements and identify every line that represents an allocation rather than a direct cost.

Then rebuild each function from the bottom up. How many finance people does this business need? What will it pay for insurance at its own scale and loss history? What will its ERP cost, licensed for its own user count? What does its benefits program cost without the parent's population? What does it pay for audit, legal, and tax? In nearly every case the answer exceeds the allocation, because the business loses the parent's scale.

Build the three-number bridge: allocated cost in the statements → TSA cost during transition → standalone cost afterward. Deals are mispriced when the buyer models the first number and pays the second.

Separately model one-time separation costs: IT implementation, rebranding, recruiting, facility fit-out, legal and advisory fees, and duplicate running costs during transition. These are real cash and they are not in the carve-out statements.

On the sell side, run the mirror analysis. Stranded costs: what portion of shared services the business absorbed, what can be eliminated, on what timetable, and at what one-time cost. Do this before announcing the divestiture's financial benefit.

Phase 6: Negotiate the TSA

Have the service owners draft the schedules. Not lawyers, not the corporate development team — the people who actually perform the service. Vague scope is the largest single source of TSA disputes, and the cure is a description written by someone who could perform the service from it.

One schedule per service, containing: description; who performs; volume assumptions and what happens outside them; dependencies on the buyer; systems used; the seller personnel involved; price and pricing basis; duration and extension rights; service level and measurement; and exit deliverables.

Price deliberately. Cost-plus, fixed fee, or pass-through, with third-party costs passed through at actual. Escalate over time — cost plus five percent initially, rising materially after months twelve and eighteen — because the escalation is what motivates exit. Buyers should accept escalation in exchange for tighter service levels and clearer exit obligations.

Set duration by service, with IT longest. Extensions on notice, at a premium, with an outside date beyond which no extension is available.

Benchmark service levels to the seller's own historical internal performance in the twelve months before closing — not to a market SLA the seller never met. Remedies are typically service credits; the liability cap is typically a multiple of fees. The buyer should understand that a TSA failure costing millions may yield a remedy measured in thousands, and should protect itself through migration speed rather than through remedies.

Negotiate the exit hardest. Termination for convenience by the buyer on short notice without penalty. A migration plan with milestones and mutual cooperation obligations. Data extraction in a specified, usable format, with a deadline. Knowledge transfer and reasonable access to the seller's personnel. And data deletion obligations, certified, on both sides.

Do not let the TSA become management. Defined services with defined interfaces, not general operation of the business.

Ask what the retained business needs from the divested one. Every time. The reverse TSA — and the undocumented intercompany supply arrangement behind it — is the item most often discovered three weeks before signing, and it is pure leverage for whoever finds it first.

Phase 7: IT separation

The longest workstream, and the one that sets the TSA's term.

Choose the approach early.

Clone and divide. Copy the whole environment, then each side deletes what is not theirs. Fast to stand up; expensive in licensing; and each party holds the other's data until deletion is verified — a data protection obligation, not just a commercial one.

Extract and implement. The buyer builds a new environment and migrates extracted data. The usual choice for a business of any size, and the reason IT TSAs run eighteen to twenty-four months.

Continue on the seller's systems. Postpones the problem and entangles the parties. Occasionally right for a very small carve-out; usually regretted.

Handle the license consequences. Enterprise agreements typically permit use by the licensee's affiliates, and the divested business stops being one at closing. Continued use during the TSA requires the vendor's agreement, and vendors know their position. Budget for TSA-period license repricing — it is predictable and it is routinely omitted from models.

Run the full inventory, not just ERP: email and domains; network separation at shared sites; identity and access management; endpoint management; security tooling and monitoring; backup and disaster recovery; data warehouse and analytics; product-embedded software and its update infrastructure; customer portals; EDI connections with customers and suppliers; and telephony.

Two rules. The separation approach and duration should be known before signing, because they price the TSA. And data deletion should be explicit, deadlined, and certified on both sides, or each party will hold the other's data indefinitely.

Phase 8: Regulatory, permits, and registrations

Start on day one. These timelines routinely exceed the deal timetable and cannot be compressed.

Antitrust. Hart-Scott-Rodino notification, 15 U.S.C. § 18a, applies to the carve-out perimeter like any transaction. Where the divestiture is itself a remedy in another matter, expect the agency to scrutinize whether the business is viable standalone — putting the TSA, the shared contracts, and the employee list in front of a regulator.

Sector permits and licenses. Environmental permits, FDA registrations and establishment listings, export licenses and registrations, financial services authorizations, transport and hazardous materials authorizations, and state licenses. Each has its own transfer or reissuance process; several have no transfer mechanism at all and require a fresh application.

Government contracts. Novation is a formal process with the contracting agency, and it takes as long as it takes. Plan for performance under the existing contract during the interim.

Product registrations and certifications. Certifications issued in the parent's name — quality system certifications, product safety marks, customer approvals — must be reissued to the new entity. Customer requalification of a supplier whose legal identity changed can take months, and it can interrupt revenue.

Data protection. A lawful basis for transferring personal data to the new controller, notice obligations, and a transfer mechanism for cross-border flows. Update records of processing, and address the shared-systems period in a data processing agreement under the TSA.

Phase 9: Day-one readiness

Day one is the test of everything. Run it as a checklist, function by function, with a named owner and a green/amber/red status reviewed weekly for the final eight weeks.

Legal. Entities formed and in good standing in every jurisdiction. Board and officer appointments. Signature authority. Contracts assigned or in their alternative arrangement. Regulatory approvals obtained or interim arrangements documented. Insurance bound in the new entity's name.

Finance. Bank accounts opened and funded. Payment and receipt capability. Chart of accounts. Opening balance sheet. Tax registrations. Payroll capability — through the TSA if not standalone. Ability to invoice customers and to pay suppliers on day one.

People. Offers accepted and start dates confirmed. Payroll set up. Benefits effective. Access to systems provisioned. Leadership in place.

IT. Email working. Network access. Core applications available. Help desk answering. Cybersecurity coverage continuous through the transition — the gap between the seller's monitoring stopping and the buyer's starting is a genuine risk.

Operations. Materials flowing. Production running. Logistics arranged. Quality systems documented in the new entity's name.

Commercial. Customers notified. Purchase orders redirected. Pricing and terms confirmed. Sales force enabled to sell and to quote.

Communications. Employee, customer, supplier, and regulator communications prepared and sequenced.

The standard for day one is that nobody outside the company notices anything. Achieving it requires the checklist to be run to closure, not to the closing date.

Phase 10: The first two years

Months 1–3: stabilization. Fix what broke. Run the TSA governance meeting weekly. Log issues with owners and dates. Start the migration workstreams immediately — the most common failure is treating month one as a period of rest.

Months 3–12: migration. Execute the IT implementation, hire the standalone functions, replace the pass-through contracts, complete the rebranding, and exit TSA services one by one as each becomes unnecessary. Exit services individually rather than waiting for a global cutover; every service exited reduces cost and reduces dependence.

Months 12–24: exit. Pricing has escalated; the seller's people who performed the services have been reassigned or have left; and the buyer's own capability should now exist. Complete data extraction, verify deletion on both sides, and terminate.

Run the wrong-pockets provision actively. Assets and liabilities on the wrong side surface for a year or more. Keep a running list and transfer them in batches rather than one at a time.

Expect the extension request. The IT implementation will be late; almost all of them are. The seller should grant an extension at a premium in exchange for a milestone-based migration plan, and the buyer should ask before it is desperate.

Two relationship points. First, the seller's incentive to perform well decays continuously, which argues for exit speed over service level negotiation. Second, the people who know how the divested business worked inside the parent will leave the seller within a year — capture what they know while they are still there.

Worked example: Halloran Industrial's filtration divestiture

Halloran Industrial sells its filtration business — six hundred million in revenue, four plants across the United States, Germany, and Malaysia, eleven hundred employees — to a private equity buyer building a standalone platform.

Months −6 to −3: seller readiness. Halloran produces carve-out financials, models stranded costs at 31 million absorbed with 19 million eliminable, inventories 241 shared contracts, and — critically — discovers that a 2011 Japanese technology license prohibits transfer "including by operation of law or by any change in control of the licensee." That single finding changes the structure conversation entirely, because the planned reorganization would breach it under the reasoning of Cincom and PPG.

Month −2: structure. Reorganization-then-equity-sale is retained, but the license consent is negotiated before signing rather than as a closing condition. The licensor prices it at three years of royalties plus a rate increase. Halloran pays; there is no alternative.

Months −2 to 0: contract triage. 158 commodity supply agreements go to terminate-and-replace, with a quantified four-million-per-year pricing dis-synergy that goes straight into the buyer's model. 41 customer contracts go to partial assignment; 60% consent in time and the rest run through a pass-through with a two-year sunset. 29 software and technology licenses are reviewed individually — nine need consent and four licensors reprice. 13 facility arrangements at shared sites become subleases and shared facilities agreements.

Months −3 to 0: people. The list negotiation takes six weeks. In Germany, works council consultation becomes the critical path — not antitrust clearance, which everyone had assumed. In the United States, WARN analysis under 29 U.S.C. § 2101 is run before the offer list is fixed; health continuation under 29 U.S.C. § 1161 for the non-hired is allocated to Halloran; and the 401(k) spin-off is documented under 26 U.S.C. § 414.

The standalone cost model. Carve-out EBITDA of 78 million after 31 million of allocated cost. The buyer's bottom-up rebuild puts the true standalone cost at 43 million. The twelve-million gap moves enterprise value by roughly nine times that amount and is the most consequential number in the diligence.

The TSA. Thirty-one services, seven functions. IT at twenty-four months because extracting filtration's data from a shared SAP instance requires a full implementation. Cost plus five percent, escalating to plus fifteen at month twelve and plus thirty at month eighteen. Service levels benchmarked to Halloran's own prior-year internal performance. Liability capped at twelve months' fees.

The reverse TSA. Halloran's coatings division has bought a filtration component from the Malaysian plant for nine years at an internal transfer price, under no contract at all. The buyer finds this three weeks before signing and negotiates a three-year supply agreement forty percent above the historical price. The best trade in the deal, available only because someone asked.

Day one. Payroll runs on Halloran's system under the TSA. Customers are invoiced from Halloran's SAP with remittance to a new lockbox. Products still bear the Halloran name under an eighteen-month transitional license with quality control provisions. Nobody outside notices anything.

Month fourteen. The SAP implementation is four months late. TSA pricing has escalated twice. The buyer requests an extension; Halloran grants it at a further premium and requires a milestone-based migration plan. This is the normal outcome, and the parties who wrote milestones into the exit provisions are the ones not arguing about it.

Month twenty-two. Final TSA service exited. Data extraction complete, deletion certified both ways. Rebranding finished four months earlier. Of the 241 shared contracts, six remain in pass-through arrangements and are the subject of a wrong-pockets true-up.

Carve-out financial statements, practically

What they are. The business's revenues, costs, assets, and liabilities extracted from the parent's consolidated records, with allocations for everything shared.

Preparing them. Decide the perimeter first — the statements must describe the same business the agreement sells. Then: direct revenue and cost by legal entity and cost center; allocated shared costs with a stated methodology (headcount, revenue, transaction volume, or square footage, disclosed); intercompany transactions priced as they were, with the pricing disclosed; a carve-out balance sheet including an allocation of shared assets and liabilities; and parent net investment in place of equity, because the business has none.

What the buyer's quality of earnings work will test. The allocation methodology and its reasonableness. Completeness — whether costs the business consumed were captured at all. Separation and one-time costs excluded from adjusted EBITDA. The standalone estimates. Working capital normalization for a business that used the parent's balance sheet, including receivables collected centrally and payables paid centrally. And intercompany transactions not priced at arm's length.

Where the seller gets challenged most. Understated corporate cost allocation, which flatters EBITDA. Pro forma standalone estimates that assume the business will run leaner than any comparable standalone company does. And working capital that looks light because the parent financed it.

Audit. If the transaction requires audited carve-out statements — because of financing or securities requirements — build the timeline around the audit, which for a first-time carve-out audit can take four to six months and will surface allocation questions nobody anticipated.

The number that actually matters. Not carve-out EBITDA. The bridge from carve-out EBITDA to standalone EBITDA, and both sides should build it, because whoever builds it better wins the negotiation over price.

What changes with the buyer

A strategic buyer with its own infrastructure wants a short TSA and a fast migration onto its own systems, and may need no finance, HR, or IT services beyond a few months. Its diligence concentrates on the perimeter and on whether customers and technology travel intact.

A financial buyer building a standalone platform needs everything, and its TSA is the most important document in the transaction. The standalone cost model is its central diligence exercise, and its first-year priority is hiring a management team and a back office that do not exist.

A competitor buyer raises antitrust review, requires clean team protocols, and makes the cross-licenses genuinely adversarial.

A management buyout brings financing constraints and a specific optimism problem: the management team has run the business inside the parent for years and sincerely underestimates what the parent was providing. The standalone cost conversation is hardest with them, and it is the one most worth having.

And the seller's identity matters too. A public company divesting for portfolio reasons has a stranded cost and guidance problem. A sponsor exiting a platform wants a clean break and short TSA. A distressed seller may be unable to perform the TSA at all, which makes seller creditworthiness a live diligence item — and argues for shorter TSA terms, prepayment or escrow of TSA fees, and step-in rights the buyer can actually exercise.

Real estate and shared sites

Where the two businesses share a physical site, contract drafting can allocate the risk but cannot make the arrangement comfortable.

Configurations. A building housing both businesses. A campus where one owns the land and the other occupies part. A plant where the divested line sits inside the retained facility, sharing utilities, receiving, and a single environmental permit.

Instruments. A sublease or lease for the divested business's space, with a term matched to the buyer's relocation plan. A shared facilities agreement covering utilities, security, receiving, waste, maintenance, and parking, with cost allocation and service standards. Access and easement rights where crossing retained land or using retained utilities is necessary.

Environmental. A shared permit generally cannot be split; one party remains the holder and the other operates under an arrangement the regulator knows about. Allocate historical contamination separately from operational responsibility, and take a baseline assessment at closing.

Relocation. If the buyer will move out, set a target date, allocate the cost, and specify restoration. If the buyer will stay, document a real lease with real terms rather than a transitional accommodation that becomes permanent by default.

Third-party landlords. Every one of these arrangements needs the landlord's consent where the site is leased, and a landlord aware of the timetable behaves accordingly. Start early, and budget for the consent fee.

Cross-border complications

Every jurisdiction is a separate project. A carve-out spanning six countries is six transfers, six sets of local counsel, six employment regimes, six tax analyses, and six sets of permits and registrations.

Employment. Automatic transfer regimes in Europe and elsewhere move employees with the business by operation of law, with information and consultation obligations attaching before completion. The consultation period is not compressible and frequently determines the closing date. Works council agreement is required in some jurisdictions for some matters.

Tax. Transfer taxes, VAT, stamp duties, capital gains on intercompany transfers, withholding on cross-border payments, and permanent establishment questions when the seller performs TSA services into another jurisdiction. TSA pricing itself raises transfer pricing questions where the parties are related for part of the period.

Corporate. Local entity formation timelines vary from days to months. Directors and local officers may need to be residents. Capitalization requirements, registration, and bank account opening — bank account opening in some jurisdictions takes longer than everything else combined.

Regulatory. Licenses, permits, and registrations rarely transfer; several require the new entity to apply afresh with no shortcut.

Currency and treasury. Cash pooling and intercompany funding arrangements terminate at closing; the new entities need their own funding, their own accounts, and their own hedging.

Data. Cross-border transfer mechanisms for personal data, both for the transfer itself and for the shared-systems period under the TSA.

The practical rule. Sequence the jurisdictions by their longest lead time, not by their revenue. The smallest country with the slowest permit office sets the closing date.

Running the separation management office

The structure decides more outcomes than any single document, and it is worth building properly.

Composition. A full-time separation lead on each side, empowered to decide. Functional workstream leads for IT, finance, HR, legal, tax, procurement, operations, real estate, quality and regulatory, and communications. A program manager who owns the integrated plan. A steering committee that meets weekly and includes someone senior enough to break ties.

The integrated plan. One plan, with dependencies mapped. The dependency map is the deliverable that matters, because the critical path in a carve-out is almost never where people expect it — works council consultation, a permit reissuance, or a customer requalification will beat antitrust clearance more often than not.

Cadence. Workstream meetings weekly. Steering committee weekly for the final eight weeks and biweekly before. A single issues log with owners and dates, reviewed at every steering meeting. A day-one readiness dashboard by function, green/amber/red, from eight weeks out.

Decision discipline. Every open item has an owner, a date, and an escalation path. The most common failure is an issues log with fifty items and no dates.

Clean team protocols where the buyer is a competitor or the process is competitive: defined membership, defined permitted uses of competitively sensitive information, no dissemination to business personnel before closing, and a written protocol everyone signs.

Knowledge capture. The people who know how the divested business actually worked inside the parent will be reassigned or gone within a year. Capture process documentation, system configurations, customer histories, and undocumented workarounds while they are still there — it is the cheapest insurance in the transaction and it is almost never done.

Communications sequencing. Employees before customers; customers before suppliers; regulators on their own timetable. Prepare the materials in advance and decide who says what to whom. Announcement day is when the recruiters call, so retention must already be in place.

The purchase agreement provisions to insist on

Wrong pockets. For twelve to twenty-four months after closing, either party may identify an asset or liability on the wrong side and require its transfer for no additional consideration. Every carve-out needs one — every carve-out has misallocated items.

Specific further assurances. A general clause is inadequate. Require named cooperation on consents, novations, regulatory transfers, customer requalification, and data migration, with contacts and response times.

Non-obtained consents. The seller holds the benefit for the buyer, performs at the buyer's direction and cost, and both cooperate on replacements — with a sunset and a price consequence only for specifically identified material contracts.

Shared contract schedule categorizing every agreement by treatment path, owner, and target date.

Intercompany clean-up. All balances settled or eliminated at closing; all agreements terminated except those expressly continued; and all informal intercompany supply arrangements identified and documented — the undocumented ones are the reverse TSA nobody planned.

Insurance. Access to the seller's occurrence-based policies for pre-closing claims relating to the business, expressly granted; tail coverage where policies are claims-made. Neither is automatic.

Non-compete and non-solicit, both ways, with carve-outs for existing businesses, de minimis holdings, incidental competing operations acquired later, and general advertising.

Financing cooperation at the buyer's cost, with express limits on the seller's liability.

Purchase price adjustment built on the carve-out methodology. The working capital target must be computed the same way the closing statement will be — and the intercompany balances eliminated at closing must be treated consistently in both, which is the most common source of carve-out true-up disputes.

Environmental baseline. Where sites are shared or transferred, a baseline assessment at closing allocates historical from operational responsibility and is worth its cost.

Common mistakes

Starting separation planning after signing. The seller loses control of the TSA's length and the buyer loses the ability to price it.

Choosing the structure before reviewing inbound licenses. The reorganization step itself can breach them.

Defining the perimeter with a standard instead of a schedule. Every shared asset becomes a dispute.

Diligencing consents after announcement. Counterparties price to the timetable.

Omitting goodwill from a trademark assignment. 15 U.S.C. § 1060 requires it.

Modeling allocated cost instead of standalone cost. The gap is large and always in the same direction.

Ignoring stranded costs on the sell side. Reported margin improves; actual margin does not.

Letting lawyers draft the TSA schedules. Have the service owners do it.

Negotiating service levels harder than exit provisions. The remedy is service credits; the protection is speed.

Forgetting the reverse TSA. Ask what the retained business receives from the divested one, early.

Starting regulatory transfers late. Novations, permit reissuance, and customer requalification do not compress.

Treating month one after closing as a rest period. Migration starts on day one or the extension request is inevitable.

Practice pointers

Do phase 0 before you run the process. Perimeter, financials, standalone estimate, stranded costs, contract inventory, IT approach, separation office, retention.

Review every material inbound license for transfer language before choosing the structure, and negotiate consents before signing rather than as closing conditions.

Triage contracts into four buckets on a schedule, with owners and dates, and quantify the replacement pricing dis-synergy for bucket four.

Get the IP formalities right — 35 U.S.C. § 261, 15 U.S.C. § 1060, 17 U.S.C. § 204 — and build all three IP schedules, including the license-back.

Run the WARN and consultation analyses before the employee list is final, and expect works council consultation to be the critical path in Europe.

Build the three-number bridge — allocation, TSA cost, standalone cost — and negotiate price against the third number.

Escalate TSA pricing on purpose, and pair it with milestones, an outside date, and certified data deletion.

Ask what the retained business gets from the divested one before the other side does.

Run day-one readiness to closure, function by function, with the standard that nobody outside the company notices anything.

Exit TSA services individually as they become unnecessary, rather than waiting for a global cutover.

Related documents

This guide is general information, not legal advice, and does not create an attorney-client relationship.