Working Capital and Quality of Earnings: Where Deal Value Actually Moves

Why the peg exists at all

Most deals are agreed on a cash-free, debt-free basis: the buyer values the operating business, and cash and debt are settled separately at completion. That leaves a gap. A business can be handed over with its receivables collected and its payables stretched — technically the same business, materially less working capital, and therefore an immediate cash requirement for the buyer.

The working capital peg closes that gap. The parties agree a normalised level of working capital the business should be delivered with, and the price adjusts pound for pound against the actual level at completion. Deliver above the peg and the seller is paid more; below it and the price is reduced.

This is not a rounding item. On a mid-market deal, the difference between a peg set at the twelve-month average and one set at the seasonal low can run to several million pounds.

Setting the peg: where the argument happens

The peg should represent the working capital the business genuinely needs to operate at its normal run-rate. Getting there means three modelling decisions, each contested.

First, the reference period. A twelve-month average smooths seasonality; a three-month average reflects the most recent trading. Sellers prefer whichever is lower, buyers whichever is higher, and both will argue from principle.

Second, seasonality. A business whose working capital peaks in September and troughs in February cannot have a single peg applied to a completion at either date without one side being materially advantaged. The answer is a monthly peg schedule, and it is worth building.

Third, what is in the definition. Operating working capital is receivables plus inventory less payables — but accrued bonuses, deferred revenue, capex creditors and intercompany balances are all negotiated in or out, and each carries real money.

• Model the peg as a monthly schedule where seasonality is material, not a single number.
• Exclude cash and debt — those are settled separately under the cash-free debt-free mechanic.
• Treat deferred revenue explicitly. Buyers argue it is debt-like (an obligation to deliver); sellers argue it is working capital.
• Watch for a deliberate pre-completion working capital squeeze — aggressive collection and stretched payables in the final weeks.

The completion accounts true-up

At completion the buyer pays an estimated price based on estimated working capital. Actual completion accounts are prepared afterwards — typically within 60 to 90 days — and the difference is trued up in cash.

Model this as three lines: estimated working capital at completion, the agreed peg, and the resulting adjustment. Then model the true-up separately, because the timing matters to the buyer’s cash flow and to the funding requirement at close.

The alternative mechanic is a locked box, where the accounts are fixed at a date before signing and the buyer takes economic risk from that date forward, compensated by an interest-style ticker. Locked box is common in European auctions because it gives price certainty; completion accounts are more common where the buyer wants protection against deterioration between signing and closing.

Quality of earnings: challenging the EBITDA itself

The multiple is applied to an EBITDA figure, so any change to that figure moves the price by the multiple. On an 8.0x deal, a £1m QoE adjustment moves the price by £8m. This is why quality of earnings work is the highest-leverage diligence there is, and why the seller’s adjusted EBITDA is a negotiating position rather than a fact.

A QoE exercise separates recurring, maintainable earnings from everything else. It reverses one-off gains the seller has left in, adds back genuinely non-recurring costs, normalises owner remuneration to a market rate, strips out related-party transactions that will not continue, corrects revenue recognition and cut-off errors, and pro-formas the full-year effect of acquisitions or cost actions completed part-way through the period.

The add-backs sellers propose and buyers resist are predictable: restructuring costs that recur every year, "one-off" legal fees in a litigious sector, run-rate synergies from actions not yet taken, and management add-backs for costs the business will still incur under new ownership.

• Owner remuneration normalised to a market-rate salary for the role.
• Related-party revenue or cost at non-market terms, restated or removed.
• Restructuring costs tested for recurrence — three consecutive years of "one-off" costs are recurring.
• Run-rate adjustments for actions actually completed, not planned.
• Revenue recognition and cut-off, particularly around period ends.
• Any capitalised cost that should have been expensed.

How to model both together

Build a bridge, not a number. Start from reported EBITDA, list each QoE adjustment as its own line with a stated basis, and arrive at adjusted EBITDA. Apply the multiple to that. Then bridge from enterprise value to equity value: less net debt, less debt-like items, plus or minus the working capital adjustment against the peg.

Presenting it this way does two things. It makes every contested item visible and separately negotiable, which is how the conversation actually runs. And it lets you sensitise the outcome properly — a table showing equity value across a range of adjusted EBITDA and a range of working capital delivery is far more useful to a decision-maker than a single number.

The discipline is the same as everywhere else in modelling: every line traceable to a driver someone can challenge. A QoE bridge with an unexplained "normalisation adjustment" line is the item diligence will open first.