Carve-out accounting: standing up a finance function before the TSA ends

A carve-out closes and the new company owns a P&L it has never produced by itself. The general ledger sits in the seller's ERP. The close is run by a shared services group that works for the seller. There is a controller on the org chart, and depending on the deal there may be one in the building. Between those two facts sits a finance function that exists as a service somebody else has agreed to keep providing for a while.

That while has a date on it. The date was set in the deal documents, before anyone had run a month of work inside the new company and learned what the work actually is.

What the transition services agreement is actually holding

A transition services agreement keeps the seller running the parts of the back office the new company cannot yet run for itself. In finance that usually means ERP access and the general ledger, the month-end close, accounts payable and receivable, payroll processing, and the reporting the lenders and the sponsor expect on the old calendar. M&A advisory guides put finance and accounting services commonly in the range of six to eighteen months, with ERP and accounting among the last services to switch off.

What the agreement transfers is the service. It does not transfer the people. The group closing those books is the seller's group, working under contract, and on the day the term ends they go back to closing the seller's books. Everything they know about how the acquired business actually runs goes with them.

An extension is usually negotiable. Advisory guidance describes extension pricing set above the base rate, and extension rights that often require the seller's consent. Time can be bought, at a price, from a counterparty who has already moved on.

What happens to finance when the agreement expires

The work keeps arriving on the same calendar. The people doing it stop. On the first of the month after the term ends, the reconciliations, the payables run, the close checklist, and the reporting pack are all due, and the group that used to produce them is no longer on the other end of the email.

That work has nowhere designed to land. When new volume arrives at a company that has been running for years, it arrives at a function that already exists, and the function stretches to hold it. A carve-out has no function to stretch. The work goes to the few people the new company actually has, which usually means the CFO and whatever finance leadership came across with the deal, and they carry it themselves until something is built. The first standalone close goes out because they do. That is finance leadership doing exactly what a separation asks of it. It is also an arrangement with a cost, and that cost has a shape worth knowing in advance. Where the back-office work goes after a bolt-on traces the same cost through the quarter after a deal.

How long it takes to stand up a finance function

Longer than the recruiting calendar suggests, because recruiting is the shortest part of it. Standing up a finance function is four pieces of work in sequence. Decide what the function does now that it serves one company rather than a division of a larger one. Take control of the systems, the ledger, and a chart of accounts that is yours. Write down what “done” means for every recurring stream and who signs it. Then put people in the seats and run the cycle enough times that the output can be trusted without being checked line by line.

The last piece is the one nobody can compress. A close happens twelve times a year. A team that has closed your books three times has had three chances to learn where revenue actually gets recognized and which vendor always bills late. That knowledge does not arrive with the hire. It arrives with the repetitions, and repetitions take months of calendar however the seats were filled.

Some of the work around a carve-out is not this at all. The opening balance sheet, purchase accounting, quality of earnings, and the audit are professional engagements, and they belong with the firms that perform them. The function this article is about is the recurring work that begins the day those engagements end and never stops.

What a carve-out company needs before the TSA ends

Enough of the function to close the books without the seller, run by people who have already done it more than once. Five things carry that.

• A ledger you control. Your own system, your own chart of accounts, your own access. Everything else on this list waits on it.
• A named owner against every recurring stream. Reconciliations, AP, AR and collections, payroll coordination, the reporting pack. Any line without a name against it becomes the CFO's line the first month it comes due.
• Hands enough to run the cycle every month. A design only one person can execute is a single point of failure with a monthly deadline attached to it.
• Repetitions banked while the seller is still reachable. A close run in parallel, with someone on the other end who can answer a question about a legacy account, is worth several run alone afterwards.
• A reporting pack the company produces itself. The sponsor's calendar does not move because a TSA ended, and the first month it slips is the month the question gets asked at board level.

What sits inside those lines is preparation. What sits outside them is judgment, and the sort between the two is the one every finance team already knows how to make. A stream that ends with something handed over for review can sit in a seat. A stream that ends with an approval the company has to stand behind stays with finance leadership, which is where the line falls between an offshore finance seat and your controller.

The transition services agreement has an end date. A finance function has a build time. The only one of those you control is when you start.

What we build

Large enterprises answered the standing-capacity question a long time ago. When recurring work needed to live somewhere other than on the leadership team, they built Global Capability Centers: embedded operating teams offshore holding the work and the context that comes with it. The entity, the tax and payroll infrastructure, the lease, and the multi-year build put that out of reach for a $75M company.

The same architecture, sized for the middle market, is the Mini-GCC model, and it is what Kayana builds. Seats designed around your recurring streams, running inside your close calendar and your systems, grown one at a time. One of the companies we work with grew its embedded team to 10+ seats that way. Voluntary turnover across the teams we build runs 3–5% in year one, which matters more in a carve-out than anywhere else, because the value is in the repetitions and the repetitions live in a person.

How long that takes depends on how much has to be designed, and design takes the time it takes. The end date on the agreement is simply the reason to have the conversation while there are still months on it.

Quick answers

The design principles behind an operating layer that compounds are the subject of the book. Operational Alpha →