Locked box or completion accounts. The same dispute, two different dates

Three days after a deal closed, I was asked to read an agreement I had not negotiated, along with a printout of the transfer confirmation. The owner had sold the company for PLN 100 million. PLN 94 million had arrived in his account.

One item made the difference: the working capital adjustment. The parties had set the working capital target at the twelve-month average, PLN 18 million. Closing fell on January 31, when the company’s working capital stood at PLN 12 million, because January is its weakest month of the year. The six million difference came off the price to the last zloty.

The buyer calculated exactly as the agreement said. The seller had signed off on the target six months earlier and had not checked where in the annual cycle closing would fall.

I have changed the details so that neither the company nor the parties can be identified, and I have rounded the figures. The conversation after the fact goes much the same way every time.

Where the price adjustment comes from

The price is agreed on the basis of a balance sheet as at a given date. The deal closes a few weeks or a few months later, because the company has to be examined, the financing finalized and sometimes clearance obtained from the competition authority. In the meantime the company keeps working. It buys raw materials, issues invoices, repays loans and pays salaries. The position the buyer agreed to pay for is out of date before payment is made.

Price mechanisms answer one question: what state of the company the buyer is paying for, and who bears the effect of changes up to the closing date.

The buyer values the business as if it had no cash and no debt. The market calls this amount the enterprise value and usually calculates it as a multiple of EBITDA. What reaches the seller’s account, however, is the price for the shares: the enterprise value less net debt, adjusted for the deviation of working capital from the agreed level.

Net debt means bank loans, other loans, leases and other financial liabilities, less the cash in the company. If the company has PLN 20 million of bank debt and PLN 5 million of cash, net debt is PLN 15 million. With an enterprise value of PLN 115 million and working capital at the agreed level, the owner receives PLN 100 million.

Working capital is money tied up in day-to-day operations. In the simplest terms it is inventory plus receivables from customers minus payables to suppliers, and agreements add other current operating balances to this. A company with PLN 8 million of inventory, PLN 10 million of receivables and PLN 6 million of trade payables has PLN 12 million of working capital. A going business has to have goods in the warehouse and has to bridge the time between issuing an invoice and receiving the money. That is why a normal level of working capital is part of the price, just like the machinery.

The working capital target is the figure set in the agreement against which working capital on the closing date is compared. A surplus raises the price and a shortfall lowers it, zloty for zloty.

Working capital over the year

In a seasonal company, working capital changes from month to month. This is easy to see at a manufacturer that sells in the autumn. In the summer it produces for stock, so the warehouse is full and the bank account empty. In October and November the goods go out to customers, and receivables grow on the balance sheet. In January the customers have already paid and production has not yet started. Working capital is then at its lowest and cash at its highest.

In such companies, the difference between the highest and the lowest level of working capital during the year can reach between ten and twenty percent of annual revenue. A single annual average does not describe such a company. A curve does: twelve figures, one for each month.

A fall in working capital does not vanish without trace. The money that in January is not in the warehouse or in receivables is sitting in the bank account or has reduced the working capital loan. It should come back into the price through net debt. It does come back when the agreement credits all of that cash to the seller. The agreement may, however, treat it as operating cash, an amount that has to stay in the company to finance the next season and does not reduce net debt. In that case the fall in working capital will lower the price, and the seasonal cash will not raise it. I wrote separately about minimum operating cash. In the example at the start, the whole PLN 6 million seasonal fall in working capital reduced the price with no offset in net debt.

Two mechanisms, two dates

The completion accounts mechanism sets the final price after the deal. The parties sign the agreement with a base price and with target levels for net debt and working capital. On the closing date the buyer pays a provisional price, usually calculated from an estimate of net debt and working capital that the seller prepares a few days before closing. Then one of the parties draws up a balance sheet as at the closing date, most often within sixty days. The parties compare the actual position with the targets and pay the difference one way or the other. A seller who inflated the estimate in his own favor then pays back the excess, often out of money he has already put to other uses.

A locked box works the other way round. The parties take a balance sheet as at the end of a closed month or year, calculate net debt and working capital as at that date and fix the final price for the shares. After signing, the price no longer changes. The name comes from the fact that from the balance sheet date the company is locked to the seller: everything generated in it belongs economically to the buyer, even though the seller is still the legal owner.

This has three consequences worth knowing about before a seller agrees to a locked box. The seller may not take value out of the company: dividends, bonuses for himself, fees for services from his own companies. The market calls this the ban on leakage. The list of permitted payments is set out in the agreement, and any other payment goes back to the buyer zloty for zloty. Profit earned after the balance sheet date belongs to the buyer, and in return the seller usually receives interest added to the price, calculated from the balance sheet date to closing. The risk of a weaker quarter passes to the buyer, because it has bought results that do not yet exist.

The dispute over net debt and working capital takes place under both mechanisms. Completion accounts move it to the period after closing, and a locked box moves it to the period before signing.

In Europe, a post-closing price adjustment is used in roughly every other deal. The law firm CMS analyzed 601 European transactions from 2025 on which it advised. A price adjustment was provided for in 48 percent of them, and in Central and Eastern Europe in 51 percent. By comparison, in the United States it appears in 92 percent of deals. Among European deals without such an adjustment, a locked box was chosen in 54 percent.

Where the price disappears

With completion accounts, the outcome is decided by four provisions, and none of them concerns the price itself.

The first is the target. A twelve-month average is simple to calculate, but in a seasonal company it works against the seller unless the agreement also reflects the seasonal cash surplus in the price. In the example at the start, it cost PLN 6 million.

The second is who prepares the accounts. Most often it is the buyer, because it takes over the books after closing. The seller receives a draft and a deadline for objections, usually thirty days. Whoever writes the first version sets the starting point for the dispute. When the existing management board stays on, the seller can sometimes negotiate that he will prepare the accounts himself and the buyer will raise any objections.

The third is the definition of net debt. Provisions, accrued unused leave, leases, liabilities to related parties, bonuses due for payment: each of these items may or may not be debt, depending on one paragraph of the agreement. In a mid-sized company this usually comes to several million zlotys.

The fourth is the dispute procedure. A good agreement names an independent expert, the scope of the expert’s decision and the deadline for it. Without that, a dispute over PLN 3 million can drag on for a year, and the part of the price retained at closing sits in an escrow account all that time. The size of the retention is open to negotiation. The buyer proposes an amount equal to the possible adjustment in the worst-case scenario, and it is worth doing that calculation before any number is put on the table.

With a locked box, everything happens before signing, so it looks like an ordinary part of the price negotiation. The buyer calculates net debt and working capital as at the balance sheet date and deducts every disputed item from the price. It is the same dispute, settled earlier and without an expert. In a seasonal company, a locked box does not free anyone from the calendar. The buyer will compare working capital at the balance sheet date with the normal level and adjust the price by the difference before signing.

On top of that comes interest on the price for the period up to closing. The rate is set in the agreement in a single sentence, so it is easy to overlook. With a price of PLN 100 million and five months between the balance sheet date and closing, each percentage point of the annual rate is worth about PLN 417,000.

The three-number test

Before choosing a mechanism, I calculate three things. The first is the number of days from the balance sheet date to the expected closing. If the deal needs clearance from the competition authority or a sector regulator, a quarter has to be added to the timetable. Beyond ninety days a locked box stops being a simple solution, because the buyer takes on the results of a quarter in which it was not yet running the company, and it prices that cautiously.

The second is the difference between the peak and the trough of working capital during the year. The larger it is, the more depends on how the agreement treats the target and cash, whatever the mechanism is called.

The third is the cash the company will generate between the balance sheet date and closing, compared with the interest the buyer proposes to add to the price under the locked box option.

Once these three numbers are known, the mechanism chooses itself. With a short period to closing, verified figures and a company with no seasonality, I recommend a locked box. It ends the price discussion on the day of signing and spares both sides several months of settlement work. Private equity funds also choose it when, toward the end of their life, they want to settle with their investors without holding reserves for post-deal disputes. There is one condition: the interest should cover the cash the company will generate until closing. Otherwise the difference stays with the buyer, whereas under completion accounts the same cash would have reduced net debt and come back to the seller in the price.

With strong seasonality or a long period to closing, I choose completion accounts and turn my attention to the target.

The working capital target in a seasonal company

At Capital One Advisers we calculate the target from a monthly curve. You take working capital at the end of each of the twelve months, strip out one-off events and agree that the target is the value for the month in which closing falls.

The curve works on one condition: it has to be matched by the definition of cash. If the target for January is PLN 12 million and the agreement still credits the seller with all the cash released at that time from inventory and receivables, or with the working capital loan repaid from that money, the buyer will pay for the same thing twice: once through a lower target and a second time through lower net debt. It will spend that cash anyway, right after closing, on production ahead of the season. Two solutions are consistent. The first is an annual average for working capital, with the rule that all cash above a fixed minimum raises the price. The second is a monthly curve, with the rule that the seasonal cash surplus stays in the company as operating cash.

Another company, with the same seasonality as in the example at the start. We calculated the curve month by month and wrote it into the agreement as a table. Closing fell in February, working capital then exceeded the target for that month, and the buyer paid an extra PLN 2 million on that account. Sample wording: “The Target Working Capital is the amount set out in Schedule […] for the calendar month in which the Closing Date falls. Cash up to the amount specified in the same Schedule for the relevant month constitutes Operating Cash and does not reduce Net Debt.”

Two more things in the agreement need watching, and they are rarely remembered. The first is the accounting policies used for the closing accounts, so that the buyer does not change them right after closing. The agreement sets their order of priority: first the specific rules set out in a schedule, for example the policies for write-downs of obsolete inventory and overdue receivables, then the company’s existing practice as applied in calculating the target, and only last the general accounting rules. On top of that comes a list of the ledger accounts allocated to working capital and to net debt.

The second is the obligation to run the company in the ordinary course until closing, including paying liabilities on their usual terms. Without it, a seller whose agreement does not count cash below the minimum toward the price can pay suppliers early in December and turn that cash into working capital that will raise the price.

The same principle applies to an additional payment that depends on the company’s future results (an earn-out): the money is decided by the definition of the metric and by who calculates it. I wrote about this separately.

What the choice of mechanism will not fix

If the company’s data is weak, no mechanism will help. A locked box requires a balance sheet the buyer will believe, so nobody will propose one when the books are kept haphazardly. With books like that, completion accounts turn into a three-month dispute in which the seller has to defend items he cannot document.

The order is therefore the reverse of what most sellers assume. First a review of the books, cleaning one-off events out of earnings and a tidy balance sheet, then the discussion about the mechanism. I wrote separately about what the buyer’s analyst finds and how much each item costs when the seller has not calculated it first.

I once signed an agreement myself with an annual average as the target. Closing fell in June, it worked out well for us, and for a year I thought it had been good work. It was luck.

At Capital One Advisers we prepare the working capital curve, the operating cash minimum and the list of net debt items before the memorandum is sent out. The discussion about definitions then takes place while the seller still has several bidders.

Buyers’ advisers will say that completion accounts are fairer, because the buyer pays for the actual position. Sellers’ advisers will reply that fairness depends on who prepares the accounts and who holds the money while the dispute goes on. For a seasonal company I choose completion accounts with a monthly curve. The buyer’s adviser has a good answer to that: twelve figures mean twelve chances for a dispute over what was a one-off event, while a locked box closes the subject on the day of signing. Which of these uncertainties costs more depends on the quality of the seller’s data and on how many months separate the balance sheet date from closing.

Service: M&A advisory on a company sale