Clawback and fee retention

Three mechanisms, routinely confused

People use clawback, retention and earn-out as though they were the same thing. They are not, and a heads of terms that mixes them produces an argument later about money that has already moved.

• Retention. Money the buyer holds back, or that sits with solicitors, and is released only if a test is met. The buyer already has the cash.
• Clawback. Money the buyer has already paid and is entitled to recover if a test is failed. The seller has the cash and the buyer has a claim.
• Earn-out. Additional money that becomes payable if the business performs above an agreed level. It is upside, not protection.

The difference matters enormously in practice, because a retention is a number you subtract and a clawback is a lawsuit you may or may not win. Where the choice is available, a buyer wants the protection sitting on their side of the ledger.

The parent page on buying an accountancy practice notes that a clawback which cannot recover anything is decoration. This page is about how to build one that bites.

The base schedule is the whole thing

Every mechanism starts from an agreed list of clients and fees at completion. It should be a signed schedule attached to the agreement, not a reference to "the fees of the business", and it needs a fee figure per client rather than a total.

Fix the basis explicitly. Recurring fees only, or all fees? Gross of disbursements or net? Annualised, or the amount actually invoiced in the twelve months before completion? A client billed twice in the year before completion because two sets of accounts were caught up is not worth twice as much going forward, and if the schedule says it is, the buyer is paying for it and testing against it.

Get the schedule signed on the day rather than reconstructed afterwards. Reconstructing it later is how a straightforward test turns into an argument about which document was the real one.

What counts as a client lost

This is the definition that decides the money, and it is usually written in one line when it needs five. A workable clause deals with each of these:

• Resignation or transfer away. The obvious case, and the only one many clauses cover.
• Fee reduction. A client who stays but halves their fee. Measured by client count, that is a full retention. Measured by fee value, it is a half loss. Only one of those reflects what the buyer bought.
• Silence. A client who has not been billed within a stated period. Without this, a client who has quietly stopped using the firm is counted as retained indefinitely.
• Ceasing to trade. A client company that closes, is sold or is wound up. Reasonable clauses treat this as a loss, because the buyer has lost the fee either way, though sellers often argue it.
• Death, retirement or illness of the client, which is a real feature of an older client base and should be dealt with rather than left to be argued about at the time.

Then deal with attribution. A seller will reasonably want carve-outs for losses the buyer caused: a fee increase imposed by the buyer, a service level that dropped, a client the buyer chose to resign. Those carve-outs are fair and they are also the escape route in every disputed clawback, so define them tightly. "A fee increase above the rate the seller would ordinarily have applied" invites a debate; a stated cap on increases during the measurement period does not.

Measuring by value, and measuring more than once

Measure by fee value, not by client numbers. The two tests diverge exactly when it matters, which is when the largest clients are the ones that left, and a buyer who agreed a client-count test has agreed to be protected against the losses that hurt least.

Then think about when the test runs. A single test at the first anniversary is simple and it is also gameable, because attrition in a professional practice often follows the compliance cycle rather than the calendar. Clients tend to leave when the next piece of work falls due and they notice the letterhead has changed, which for many of them is once a year, and not necessarily in the first twelve months.

A two-stage test, at the first and second anniversaries, with the second carrying a smaller weight, is a common and sensible shape. So is a rolling measurement with a final reconciliation. What matters is that the measurement period outlasts the risk rather than the seller's patience.

The formula, and the trap inside it

Two approaches are in general use. In a pound-for-pound clause, the price falls by the lost fee itself. In a multiple-based clause, it falls by the lost fee multiplied by the same factor used to set the price, so the adjustment matches how the fee was valued in the first place.

Sellers prefer the first, buyers the second, and the difference is not small. A price built on a multiple of recurring fees that is corrected only pound-for-pound leaves the buyer carrying most of the loss on a fee they paid several times over.

Worked exampleSuppose a block of fees is agreed at one times recurring fees, and 40,000 of fees are lost in the measurement period. A pound-for-pound clause reduces the price by 40,000. A multiple clause at the same one times reduces it by the same 40,000, because the multiple is one. Change the price basis to one and a quarter times and the two clauses diverge: 40,000 against 50,000, on a single year's attrition. Neither figure is a market rate, and the point is only that the clause and the price basis have to agree with each other.

Whether the money can actually be recovered

A clause is worth what it can recover. That comes down to three questions.

Is there money left to take it from? If the deferred instalments have all been paid before the second measurement date, the buyer is suing rather than deducting. Align the payment schedule with the measurement schedule so that a payment always falls due after each test.

Is there a right of set-off? The agreement should say plainly that the buyer may deduct a clawback from any sum otherwise payable. Without it, the obligation to pay the instalment and the claim for the clawback are two separate things, and only one of them is easy.

Is there security behind it? Options include a sum retained by solicitors, a charge, a guarantee from the seller personally where they are selling through a company, or simply a longer deferral. Each has a cost to the seller and they will price it, which is why this belongs in the heads of terms rather than in the first draft of the agreement.

Who provides the numbers

The buyer runs the practice during the measurement period, so the buyer holds the data the test depends on. Sellers know this, and a clause that gives them no visibility invites suspicion at exactly the wrong moment.

Deal with it in advance: a stated form of report at each measurement date, a right for the seller to inspect the underlying records, a short window to dispute, and an independent third party to determine the point if the dispute stands. That last clause is rarely used and is worth its space, because its existence is what stops most disputes becoming expensive.

Why a lender reads the clause before it reads the price

To a funder, the clawback is the main piece of evidence about who carries the attrition risk. A deal where the seller keeps meaningful exposure for two years is a different credit proposition from one where the seller is paid in full at completion, even at an identical price.

Expect the clause to influence the structure directly. A strong, well-secured clawback tends to support a longer term and more comfort on the size of the facility. A weak one pushes the other way: less debt, more of the buyer's own money, or a shorter term. See equity contribution and vendor finance, since the deferred consideration and the clawback are usually the same money viewed from different ends.

It also affects the covenants. Where retention is the main risk, a lender may want reporting on fees retained rather than only on profit, which is a sensible thing to know is coming before it appears in a facility letter. See what lenders look for.

Where to settle it

All of the above is cheaper to agree in the heads of terms than in the first draft of the purchase agreement, and far cheaper than in month fourteen. The measurement basis, the definition of a loss, the dates, the formula and the set-off right are five lines that save five arguments. See heads of terms and buying a block of fees.

None of this replaces your solicitor, who drafts the clause, or your own professional judgement about the practice. What we add is the funding side of it. We know this market and we know specific people who have funded practice purchases before and who will tell you straight away whether the retention you have agreed supports the debt you want. Tell us how the deal is shaped and we come back to you with who can fund it and what they will want to see.

We introduce buyers who are acquiring through a limited company or a NewCo. If you are buying personally, as a sole trader or as a partnership, we will point you to an authorised firm directly instead.