Earn-outs, escrow and warranties

Why deals are not simply paid in cash

A buyer completing a transaction knows less than the seller and is paying for performance that has not happened yet. Three uncertainties follow: is the business as represented, will it keep performing, and is there something nobody has mentioned?

Cash at completion transfers all three risks to the buyer. Deferred structures share them. That is the entire logic, and it explains why a seller demanding all cash with no warranties will generally receive a lower price — they are asking the buyer to price every unknown into the number itself.

The practical consequence is that headline price and structure trade against each other. A higher price with more deferred and a longer liability tail can easily be worth less than a lower price paid in full at completion, and comparing offers on the headline alone is one of the more expensive mistakes a seller can make.

Earn-outs, and how they fail

An earn-out defers part of the consideration, contingent on the business hitting agreed targets after completion. It bridges genuine disagreement about the future: the seller believes performance will continue, the buyer is not certain, so they agree to settle it with evidence.

In principle elegant. In practice earn-outs are the most disputed provision in M&A, and the failure modes are consistent.

The seller no longer controls the business. Targets depend on decisions the buyer now makes — pricing, investment, staffing, which customers to pursue. A buyer can act entirely reasonably and still make the earn-out unreachable.

Allocation is contestable. If the business is integrated, whose revenue is whose? Group overheads allocated to the acquired unit reduce its measured profit, and there is usually no clean answer.

The metric can be gamed on both sides. Revenue targets encourage the seller to chase unprofitable sales; profit targets encourage the buyer to load costs.

Four things make them work better: a short period, ideally twelve months and rarely beyond twenty-four; a metric close to the top line, which is harder to manipulate than profit; explicit protections about how the business will be run during the period; and defined access to the information needed to verify the calculation. And a working assumption worth holding: treat the earn-out as a possible bonus, not as part of the price.

Escrow and holdbacks

Escrow retains part of the consideration with a third party for a defined period, available to the buyer if warranty claims arise. A holdback does the same thing with the buyer retaining the funds directly — less protective for the seller, since release depends on the counterparty.

Typical amounts run around ten to twenty percent of consideration for twelve to twenty-four months, with variation by risk. It is the buyer's practical remedy: without it, enforcing a warranty claim means suing a seller who has already been paid and may have distributed the proceeds.

Two provisions are worth negotiating. Staged release — half at twelve months, the remainder at twenty-four — rather than everything at the end. And a clear claims mechanism defining what constitutes a claim and how disputes are resolved, so funds are not held indefinitely against a vague assertion.

Warranties and indemnities

A warranty is a statement of fact about the business: the accounts are accurate, there is no undisclosed litigation, the company owns its intellectual property, tax returns have been filed. If a warranty turns out to be untrue and the buyer suffers loss, they can claim.

An indemnity is a promise to cover a specific identified risk, pound for pound, without the buyer needing to prove breach or quantify loss in the same way. Indemnities are used for known exposures — a live tax query, pending litigation — and they are a much stronger remedy.

The disclosure letter is the seller's principal defence. It qualifies the warranties by setting out what is actually true, and a buyer cannot claim for something properly disclosed. Time spent on disclosure is time spent buying protection, and it is routinely under-resourced in the final weeks when everyone is tired.

Caps, baskets and time limits

Seller liability is bounded by three mechanisms, and all three are negotiable.

A cap limits total liability, frequently to a percentage of consideration — often between twenty and one hundred percent for general warranties, with tax and title warranties commonly capped higher or not at all.

A basket or de minimis sets a threshold below which claims cannot be brought, preventing a stream of trivial claims. It matters whether the basket is a true excess, where only the amount above it is recoverable, or a tipping basket, where crossing the threshold makes the whole amount claimable.

Time limits restrict how long claims may be brought — commonly twelve to twenty-four months for general warranties, longer for tax, reflecting statutory assessment periods.

Increasingly, warranty and indemnity insurance is used to transfer this exposure to an insurer, allowing a cleaner break for the seller and a solvent counterparty for the buyer. It costs a premium and is well established in larger transactions; it is becoming more common in African mid-market deals where sellers want genuine finality.

Comparing two offers properly

Reduce both to the same terms. What is paid in cash at completion? What is deferred, over how long, and what has to happen for it to be paid? How much sits in escrow and for how long? What is the warranty cap and the claims period? What is the realistic probability, in your own judgement, of the deferred amount actually being received?

An offer of R100m with R60m at completion, R25m in a two-year earn-out and R15m in escrow is not an offer of R100m. It is R60m certain plus two contingent amounts, and it may well be worth less than a clean R80m. Sellers who compare headline numbers regularly choose the worse deal.