Carve-outs: building standalone financials a buyer can underwrite

Non-core disposals lead the 2026 sell-side pipeline. In carve-outs, value is lost in the gap between allocated and standalone costs, in TSA pricing and in a working capital nobody has measured on its own.

Non-core disposals by larger groups came out as the first expected sell-side driver in the CMS European M&A Outlook 2026, cited by 42% of respondents, ahead of distressed sales and private equity exits. KPMG's 2026 outlook went further and called 2026 "the year of the carve-out". What we see in our own pipeline is consistent: a growing share of the autumn processes are divisions of French and European groups, often in business services and healthcare, being prepared for sale over the summer.

These assets rarely fail on the business itself. They lose value in three places: the standalone cost base, the transitional services agreement (TSA), and a working capital that has never been measured outside the group. This note is the checklist we use when we prepare, or challenge, carve-out financials.

Where carve-out numbers come from

A division usually has no full set of accounts of its own. It may sit across several legal entities, share a payroll, a treasury and an ERP with the rest of the group, and receive a monthly management charge that nobody has reviewed in years.

IFRS has no standard on combined or carve-out financial statements. The Big 4 guides describe the same two approaches: either measure central services on a basis that reflects what the business would pay, or allocate the parent's actual costs using a reasonable key. KPMG's guide notes that a key based purely on size, such as revenue or assets, is likely to be inappropriate unless it correlates directly with the cost. PwC lists the keys that usually hold up: specific identification, legal obligation, headcount and time spent.

The first conclusion for the seller is simple. An allocation that is acceptable for historical carve-out accounts is not, by itself, a standalone cost base. A buyer underwrites what the business will cost to run on day one after the TSA, not what the group used to charge it.

1. Allocated costs versus standalone costs

We split every central cost line into three columns: what is allocated today, what the business will actually need, and when the gap appears.

Two lines are systematically underestimated. Purchasing scale does not appear in the management charge at all, because it sits in gross margin: a healthcare services division buying medical consumables under a group contract, or a facility services business using the group's fleet and temp-staff framework, will see its unit costs move when those contracts end. Insurance is the second: BakerHostetler points out that cover often reflects enterprise-wide programmes that do not translate cleanly to a standalone entity, and that deductibles and self-insured retentions need to be allocated in the SPA.

Worked example. A business services division sold at €70m EV on a reported EBITDA of €8.2m. The group charge was €1.9m. Our standalone build came to €3.1m once IT licences were repriced, a finance team of four was hired, insurance was quoted on a standalone basis and two purchasing contracts were re-tendered. The €1.2m gap, at the 8.6x mid-market multiple Argos reported for the first quarter of 2026, is worth roughly €10m of EV. Presented by the seller with quotes and an organisation chart, most of it was accepted in the price. Discovered by the buyer, it would have been taken in full, with a margin of caution on top.

Keep two notions apart, as several advisers have stressed this year. Stranded costs stay with the seller after the sale and are the seller's problem. Standalone dis-synergies are incurred by the business being sold and belong in its EBITDA. Mixing them is the most common error in the carve-out P&Ls we review.

2. TSA pricing and exit costs

The TSA bridges the period until the business has its own systems and teams. It usually covers six to twelve service categories, priced at cost or cost plus a mark-up, commonly in the 5–10% range, with step-ups if the buyer extends beyond the agreed term.

Three questions decide how the TSA plays in the price:

1. Is the TSA fee above or below the standalone cost? If the TSA charges €2.4m a year and the standalone run-rate is €3.1m, the buyer's EBITDA for the TSA period is flattered. Adjusted EBITDA should reflect the standalone cost, not the TSA fee.
2. What will exiting the TSA cost? Data migration, a new ERP instance, recruitment, dual running and re-papering of contracts are one-off costs. They are not EBITDA, but they are real cash. We present them as a separation cost estimate that the buyer can treat as debt-like or negotiate as a seller contribution.
3. Who carries the overrun? If the TSA is priced per service line, the fee steps down as each service is exited. A single global fee keeps the buyer paying for services it no longer uses. Step-up clauses for extensions should be modelled, because IT exits slip more often than not.

Two variations are worth knowing. Reverse TSAs, where the divested business continues to serve the group, create revenue that may disappear at the end of the term and must be shown separately in the quality of earnings. And some sponsors now push for carve-outs with no IT TSA at all, standing up systems before completion; this moves cost and execution risk to before signing, and the seller's timetable needs to allow for it.

3. Intra-group flows and carved-out working capital

Working capital is where carve-out numbers are least reliable, because some of it does not exist inside the perimeter.

• Intra-group trading balances. Receivables and payables with the group should be identified by counterparty and split between trading balances, which will be settled in cash, and financing balances, which are debt or equity. Balances to be forgiven are part of net parent investment, not working capital.
• Cash pooling. Sweep accounts held by the parent are excluded from the perimeter. The pool must be unwound in the legal steps plan and the business needs its own bank lines on day one, which affects the minimum cash we assume in the equity bridge.
• Centralised processes. Where the group collects customer cash, runs factoring or pays suppliers centrally, the division's historical receivables and payables are not its own. In healthcare, reimbursements collected through a group entity can distort DSO for months.
• The peg. A normalised working capital target built on group-supported balances will be wrong. We rebuild twelve to twenty-four months of monthly NWC on a standalone basis, with the same definitions as the completion accounts, and explain each reclassification.

What to do before the process starts

1. Map every central service the division receives and give each one an owner, a cost, an allocation key and a standalone estimate supported by quotes or benchmarks.
2. Price the TSA per service line, at cost or cost plus a stated mark-up, with an exit plan and dates for each line.
3. Estimate one-off separation costs and decide early whether the seller contributes or the buyer prices them as debt-like.
4. Identify purchasing contracts and insurance programmes that will not transfer, and get indicative standalone pricing.
5. Classify intra-group balances, agree the cash pooling exit and rebuild monthly standalone NWC before setting the peg.
6. Put all of this in the vendor due diligence report. A buyer will accept a standalone cost base it can audit; it will discount one it has to rebuild.

This note is general information, not advice on a specific transaction.