Deal terms
The MAC clause almost never works, and that is the point
In twenty-five years only one buyer has walked away on a material adverse change and had a court agree. Sellers should still rea…
9 November 2026 · 4 min
The median escrow on an uninsured deal is 11.1% of the price. On an insured one it is 2.8%. Most sellers still negotiate the holdback before anyone has priced a policy, which is the wrong order, and it costs them the difference.
All M&A Research
Deal structure and process
Updated
Key takeaways
Ask a founder what they expect to be held back from the price at completion and most have a number in mind, usually around ten per cent. They are right about the market and wrong about their own deal, because the number has split in two and which half you land in is largely a matter of sequencing.
In SRS Acquiom’s 2026 study of more than 2,300 private-target transactions closed between 2020 and 2025, 88% of 2025 deals included an escrow or holdback. The median across all of them was 10.0% of deal value: stable, and the same figure as the year before.1
Split that by whether the deal carried warranty insurance and the stability disappears:
| 2025 deals | Median escrow |
|---|---|
| Without warranty insurance | 11.1% |
| With warranty insurance | 2.8% |
On the indemnification escrow specifically, the pot covering warranty claims rather than the price adjustment, the gap is wider still: a median of 10.0% uninsured against 0.5% insured.1
The arithmetic is not subtle. On a £20m sale, the difference between the two medians is roughly £1.66m of your own money, held by somebody else, for eighteen months.
Uninsured, at 11.1%: £2.22m locked up until the warranty period expires.
Insured, at 2.8%: £560,000.
Against that, the premium. Marsh’s transactional risk market map put average US and Canadian rates at 3.0% to 3.7% of the policy limit as at the first quarter of 2026.2 The word doing the work there is limit. Policies are commonly written at around 10% of enterprise value, so on a £20m deal a £2m limit at 3.5% costs roughly £70,000: about 0.35% of the price, not 3.5% of it.
Even where the seller pays the entire premium, £70,000 to release £1.66m eighteen months earlier is not a close call.
Not because they reject the maths. Because of the order in which the questions arrive.
The escrow percentage is negotiated early, often in the letter of intent, when both sides are still establishing the shape of the deal. Warranty insurance gets considered later, during diligence, once the lawyers are engaged and the broker is instructed. By then the holdback has been agreed and nobody reopens it.
The sequence that works is the reverse. Get an indicative quote before any escrow is discussed, and negotiate the holdback knowing what is achievable. A seller who arrives at that conversation able to say “with a policy in place the market median is 2.8%” is negotiating from the same data as the buyer’s adviser.
Warranty insurance was identified on roughly 46% of deals in the study.1 It is no longer the unusual option that has to be justified; on a mid-market deal of any size its absence is the thing that needs explaining.
Two limits worth being clear about, because the product is sometimes sold as more than it is.
It does not cover what you have disclosed. Insurers underwrite the unknown. Anything in the data room or identified in diligence is a known matter and is excluded. This means cover narrows as the process proceeds, and that light diligence produces broad exclusions rather than cheap cover, because the insurer prices the gaps it cannot see into.
It does not cover the price adjustment. The completion accounts true-up (cash, debt and the working capital peg) is not a warranty breach. Insured deals still carry a separate adjustment escrow, typically around 1% of transaction value, and three-quarters of deals with an adjustment mechanism hold one.1 A seller asking how much is being held back needs the total across both pots, not the headline on one.
Price the policy before you negotiate the holdback. That is the whole of it, and it is a sequencing decision rather than a technical one, which is why it is so often missed by sellers who have thought carefully about everything else in their deal.
SRS Acquiom’s 2026 private-target M&A deal terms study, covering more than 2,300 transactions closed between 2020 and 2025, as summarised in Fasken, “Private M&A Deal Trends to Watch: Key Takeaways from SRS Acquiom’s 2026 Study” (7 May 2026). The sample is predominantly US buyers, so the levels are a reference point rather than a UK market rate. ↩ ↩2 ↩3 ↩4
Marsh transactional risk market map, average representations and warranties premium rates and retentions for the US and Canada as at Q1 2026, as reported in trade coverage of Marsh’s transactional risk data. Broker-published ranges differ; obtain a current quote rather than relying on any single published figure. ↩
Related
Deal terms
In twenty-five years only one buyer has walked away on a material adverse change and had a court agree. Sellers should still rea…
9 November 2026 · 4 min
Deal terms
A seller can spend six weeks arguing a warranty schedule down and then hand over a disclosure letter that gives most of it back,…
9 November 2026 · 4 min
Deal structure
Buyers want assets and sellers want shares, for the same reason: one of them leaves the history behind. Where that argument land…
9 November 2026 · 4 min