Warranties and Indemnities in a Business Sale

A warranty and an indemnity are not two ways of saying the same thing. A warranty is a contractual statement of fact about the business being sold, and breaking one gives the buyer a claim in damages measured by how much less the business was actually worth than it would have been had the statement been true. An indemnity is a promise to pay a specific sum if a specific thing happens, and if it is drafted properly it pays pound for pound without the buyer having to prove the business was worth less than the price.

That distinction decides who carries which risk on almost every share sale we handle, and it is the part of the sale agreement that most often gets glossed over until something goes wrong. We act for both buyers and sellers of owner-managed businesses, so this guide sets out the legal position as it actually works, with the arithmetic shown, rather than the familiar line that “an indemnity is stronger” with nothing behind it.

Warranty versus indemnity: the real difference

The difference is not about how serious the promise is. Both are contractual promises and both are enforceable. The difference is about what the buyer has to prove and what it gets paid.

A warranty claim is a claim for damages for breach of contract. The buyer has to establish that the statement was untrue, that it suffered loss as a result, and how much that loss was. The amount is governed by the ordinary rules of contract damages, which include remoteness and the duty to take reasonable steps to limit the loss.

An indemnity, when drafted as a covenant to pay rather than as a promise to make good a loss, operates as a claim in debt. The buyer proves that the trigger event happened and proves the amount. It does not have to show that the shares were worth less than it paid, because the indemnity is not compensating a drop in value, it is reimbursing an identified cost.

What a buyer actually recovers for breach of warranty

This is the point most published guidance either skips or gets slightly wrong, and it matters more than any other clause in the agreement.

The measure is the drop in value, not the cost of the problem

Damages for breach of contract put the claimant in the position it would have been in had the contract been performed. Applied to a warranty about the condition of a company, that produces the measure Lord Hoffmann set out in Lion Nathan Ltd v C-C Bottlers Ltd [1996] 1 WLR 1438, which the High Court quoted and applied in Triumph Controls UK Ltd v Primus International Holding Co [2019] EWHC 565 (TCC) at paragraph 488: the purchaser is prima facie entitled to “the difference between what the goods as warranted would have been worth and what they were actually worth”.

In a share sale that means the difference between the value of the shares had the warranty been true and their actual value at completion. It does not mean the cost of putting the problem right, and it does not mean the face value of the liability that turned up.

The Court of Appeal applied exactly this in Decision Inc Holdings Proprietary Ltd v Garbett [2023] EWCA Civ 1284. The warranted earnings supported an “as warranted” value of £6.43m, the true value of the company at completion was found to be £3,690,493, and the prima facie damages were therefore around £2.74m. The Commercial Court took the same approach in 116 Cardamon Ltd v MacAlister [2019] EWHC 1200 (Comm), comparing the warranted accounts with the restated position.

Two consequences follow, and they cut in opposite directions.

• Where the price was set as a multiple of earnings and the warranty breach reduces maintainable earnings, the damages can be several times the annual shortfall. A £80,000 overstatement of sustainable profit on a six times multiple is a £480,000 valuation problem.
• Where the breach is a one-off historic liability that does not affect what the business will earn in future, the damages may be far smaller than the headline figure, and in some cases nil. If the buyer struck a keen price and the shares were still worth what it paid despite the breach, there is a breach but no recoverable loss.

The buyer has to prove the loss, and that is expensive

Proving the two valuations usually means instructing a forensic accountant or valuer on each side, and the cases above were decided on competing expert evidence. On a £2m to £10m owner-managed business deal, the cost and delay of running that argument is a real deterrent to bringing a claim at all. Sellers know this. It is one reason buyers press for indemnities on anything already identified in due diligence, because an indemnity converts a valuation dispute into an arithmetic exercise.

Remoteness and the duty to mitigate

Because a warranty claim is a damages claim, the ordinary limiting rules apply. Loss that is too remote, meaning loss that was not within the parties’ reasonable contemplation at the time of the contract as a likely result of the breach, is not recoverable. The buyer must also take reasonable steps to limit its loss, and cannot recover for loss it could reasonably have avoided. Most well-drafted sale agreements restate the duty to mitigate expressly in the seller limitations schedule, so the point rarely has to be argued from first principles.

Sale agreements also normally state that the warranties are given as contractual terms only and not as representations, and exclude any remedy for misrepresentation. The purpose is to shut out a parallel claim under section 2(1) of the Misrepresentation Act 1967 and the right to rescind, both of which would sit outside the carefully negotiated caps and time limits.

What a buyer actually recovers under an indemnity

An indemnity reimburses the buyer, or sometimes the target company directly, for a defined cost. There is no valuation exercise. The buyer shows the trigger has occurred and shows what it paid out.

The trade-off is that a court will hold the parties to the words they used. In Wood v Capita Insurance Services Ltd [2017] UKSC 24 the Supreme Court construed an indemnity in a share purchase agreement which covered losses following from “claims or complaints” relating to mis-selling. The buyer’s actual loss arose from a self-reported issue and a redress exercise agreed with the regulator, not from any customer claim or complaint. The Supreme Court dismissed the buyer’s appeal, holding that the indemnity was triggered only in limited circumstances. The buyer had also failed to notify a warranty claim within the two year window, so it recovered nothing at all.

That case is the single best argument we can give a buyer for spending time on indemnity wording rather than assuming the label does the work. An indemnity is only as wide as its trigger. If the risk you are worried about can crystallise in more than one way, the trigger has to cover each of them.

There is a further point that is often stated too confidently elsewhere. It is correct in principle that a true covenant to pay is a debt claim, so remoteness and mitigation have no natural role. It does not follow that a clause labelled “indemnity” automatically escapes those rules, because the courts construe indemnities strictly and will look at what the clause actually promises. The safe course, and the one we take in drafting, is to say expressly that the indemnified sums are payable on demand, are not subject to any requirement to mitigate, and are not reduced by any rule as to remoteness.

A worked illustration: the same facts, two routes

The following example is hypothetical and is used only to show the arithmetic. Assume a buyer acquires the entire share capital of a trading company for £3,000,000, priced at six times maintainable earnings of £500,000. The agreement has a de minimis of £25,000 per claim, an aggregate basket of £100,000 recoverable on an excess only basis, and a cap of £1,500,000.

Two different problems emerge after completion. The first is a one-off employment tribunal claim relating to a pre-completion dismissal, which costs £120,000 to defend and settle. The second is that the warranted maintainable earnings were overstated: the true sustainable figure was £420,000, not £500,000.

Two lessons come out of that. For a known, quantifiable, one-off exposure, an indemnity is worth roughly six times what the same point is worth as a warranty, because the basket and the excess structure eat most of a modest warranty claim. For an unknown problem that damages the earnings the price was built on, the warranty is the valuable protection, and no indemnity is going to be offered for it.

Sellers should read the same table the other way. Agreeing a specific indemnity for something found in due diligence is agreeing to pay in full, outside the protections you negotiated for everything else.

What the warranty schedule actually covers

On an owner-managed company sale the warranty schedule typically runs to between forty and a hundred and fifty numbered statements, grouped broadly as follows. It is set out in full in the share purchase agreement, usually as a separate schedule rather than in the body of the document.

• Capacity and title. That the sellers own the shares free of encumbrances and are able to sell them. These are given without any financial cap in most deals, because a seller who does not own what it is selling has no case for limiting liability.
• Accounts and financial position. That the accounts give a true and fair view, that management accounts were properly prepared, that there is no undisclosed debt, and that there has been no material adverse change since the accounts date.
• Contracts and customers. That material contracts are listed, that none is in breach or terminable on a change of control, and that no key customer has given notice.
• Assets, stock and debtors. That the company owns its assets, that stock is useable and saleable, and that book debts will be collected in the ordinary course.
• Employees and pensions. That the employee list is accurate, that there are no outstanding claims or collective consultation obligations, and that there is no final salary scheme.
• Property. That the company holds the properties listed, that there are no arrears or breaches of covenant, and that no notices have been served.
• Intellectual property and IT. That the company owns or is licensed to use everything it needs, and that no third party rights are being infringed.
• Litigation and compliance. That there are no current or threatened claims, no regulatory investigations, and no breaches of applicable licensing or sector regulation.
• Data protection. That the company complies with UK data protection law and has had no reportable breach.
• Tax. A self-contained set covering returns, payments, elections, group arrangements and reliefs, sitting alongside the tax covenant dealt with below.

The schedule is not a formality. Every statement in it is a question the seller is being asked, and the honest answer either confirms the statement or goes into the disclosure letter. That exercise is what preparing properly for a sale is largely about, and it runs in parallel with the buyer’s own legal due diligence. In regulated sectors the schedule grows a further set of statements about licences and registrations, as our note on due diligence when buying a dental practice illustrates.

How sellers limit their exposure

Warranty liability is limited by a package of provisions, usually gathered in a limitations schedule. The individual numbers are negotiable, the structure rarely is.

Financial cap

An overall ceiling on claims, expressed as a percentage of the price. On owner-managed deals the general warranty cap commonly sits somewhere between a quarter and the whole of the consideration, with title and capacity warranties, and often the tax covenant, capped at the full price or uncapped. Where there are several sellers, each should be liable only for their own proportionate share and only for warranties they personally gave, rather than jointly and severally for the whole.

De minimis and the basket

Two separate thresholds that are frequently confused. The de minimis is a floor on individual claims: anything below it is ignored entirely. The basket, or aggregate threshold, is a floor on the total: the buyer cannot claim at all until qualifying claims added together exceed it.

The crucial drafting point is what happens once the basket is passed. On an excess only basis the seller pays only the amount above the threshold. On a first pound basis the seller pays the whole amount from the first pound. The difference between those two words can be the whole value of a modest claim. In 116 Cardamon Ltd v MacAlister the agreement carried a £500,000 threshold under which, on the buyer’s own case, the sellers became liable only for the excess, which is precisely why the parties argued about whether the shortfall exceeded it.

Time limits

Commercial warranties are typically limited to twelve to twenty four months from completion, so that at least one full audited year passes under the buyer’s ownership. Tax warranties and the tax covenant get a longer period, usually six or seven years, for the reason explained below. The clause normally requires written notice of the claim within the period, with specified detail, and then requires proceedings to be issued within a further six to twelve months or the claim falls away.

Buyers should treat those notice provisions as traps rather than formalities. In Nobahar-Cookson v The Hut Group Ltd [2016] EWCA Civ 128 the Court of Appeal held that contractual time limits for notifying claims are a form of exclusion clause, and that where the wording is genuinely ambiguous the narrower construction, meaning the one that preserves the claim, should be preferred. Briggs LJ put the underlying principle as parties not normally giving up valuable rights without making it clear that they intend to do so. That is useful if you are already in a dispute, but it is a rescue, not a plan. Diary the deadline on completion day.

Disclosure

Anything properly disclosed cannot found a warranty claim. This is the seller’s single most effective protection and is dealt with in full in our guide to disclosure letters and how sellers limit their liability.

What counts as sufficient disclosure depends entirely on the words of the agreement. In Triumph Controls UK Ltd v Primus International Holding Co the High Court drew together the principles from the earlier authorities at paragraph 335, including that a disclosure letter which purports to disclose a matter merely by pointing at another document as a source of information will generally not be fair disclosure where the contract requires sufficient detail, and that where disclosure is permitted by reference to documents outside the letter, only matters ascertainable directly from those documents are disclosed. In Infiniteland Ltd v Artisan Contracting Ltd [2005] EWCA Civ 758 the Court of Appeal held that general disclosure of everything supplied to the buyer’s reporting accountants was effective, because that was the basis of disclosure the buyer had agreed to accept.

The practical message for a buyer is to insist that the agreement requires disclosure to be fair, with sufficient detail to identify the nature and scope of the matter disclosed, and to resist a clause deeming the entire data room disclosed. The message for a seller is the mirror image, and either way the disclosure letter deserves the time it takes to do properly.

Knowledge qualifiers

Sellers ask for warranties to be given “so far as the seller is aware”. Buyers then ask for awareness to be defined as awareness after making reasonable enquiry of named individuals. The gap between those two positions is wide. Without the reasonable enquiry wording a seller can honestly say it did not know, having chosen not to look. With it, the seller is fixed with what a sensible enquiry would have turned up. Agree which individuals count, and expect the buyer to resist knowledge qualifiers entirely on the accounts, tax, title and compliance warranties.

What buyers push for in return

A buyer that has run proper due diligence will typically ask for the following, and a seller should expect to concede some of them.

• Specific indemnities for every material issue due diligence identified, outside the cap and basket, because those risks are known rather than allocated by chance.
• Warranties repeated at completion where there is a gap between exchange and completion, with a right to walk away if a repeated warranty is breached materially.
• Security for the warranty package, usually retention of part of the price in a joint account, set-off against deferred consideration or an earn-out, or a parent company guarantee.
• Carve-outs from the limitations for fraud, dishonesty and wilful concealment, which are standard and which a seller has no respectable reason to resist.
• Conduct of claims provisions, setting out who runs the defence of a third party claim that could trigger an indemnity, and on what terms.

Where part of the price is deferred or contingent, the interaction between the warranty package and the payment mechanism matters as much as the warranties themselves. A buyer with a right of set-off against future instalments has a materially better claim than one holding only a judgment against an individual who has already spent the proceeds.

Warranty and indemnity insurance, and whether it is worth it on an SME deal

Warranty and indemnity insurance is a policy covering loss arising from a breach of the warranties, and sometimes the tax covenant, in the sale agreement. Most policies are buy-side, taken out by the buyer and claimed against directly, which allows the seller’s contractual liability to be reduced to a nominal sum while the buyer still has a real covenant to claim against. Sell-side policies, which respond to the seller’s liability, are less common.

The structure is straightforward. The policy has a limit, usually a percentage of enterprise value rather than the full price, and a retention, which is the insured’s own excess and is normally aligned with the basket in the agreement. The premium is a single payment expressed as a percentage of the limit of cover, and the insurer also charges an underwriting fee and expects its legal costs to be met.

What it does not cover is as important as what it does. Policies exclude, as standard, anything the buyer actually knew about before completion, which means every issue in the disclosure letter and every risk already covered by a specific indemnity. They also commonly exclude pension underfunding, transfer pricing, secondary tax liabilities, forward-looking statements and forecasts, condition of assets and environmental liability, and they exclude fraud by the insured.

Whether it earns its place on an owner-managed deal comes down to three questions rather than to deal size alone.

• Is there a covenant problem? The clearest case for insurance is a seller who is retiring, emigrating, or is an executor or a trustee, and who genuinely cannot stand behind a warranty package for the next two years. Insurance lets the deal happen on terms a buyer can accept.
• Is the process good enough to be underwritten? Insurers will not write a policy over a deal with no proper due diligence and a thin disclosure exercise. The underwriting review sits on top of the legal work, it does not replace it, so insurance adds cost to a well-run process rather than saving it on a poorly-run one.
• Is the fixed cost proportionate? Insurers apply a minimum premium, and the underwriting fee and the insurer’s legal costs do not scale down with the deal. Below a certain transaction size those fixed costs become a meaningful percentage of the price, which is why W and I insurance is common on mid-market deals and unusual on smaller owner-managed sales, though the market for lower-value policies has broadened.

Where insurance is not proportionate, the usual alternative on a smaller deal is a retention from the price held in a joint account for twelve to twenty four months, which costs nothing and achieves much of the same comfort.

The tax covenant and how it fits

On a share sale the buyer inherits the company with its entire tax history attached, because the company is the same legal person before and after the sale. On an asset sale it generally does not, which is one of the structural differences between a share sale and an asset sale. So a share purchase agreement carries a tax covenant, historically a separate tax deed, in addition to the tax warranties.

The tax covenant is an indemnity, not a warranty. It is a covenant by the sellers to pay the buyer, pound for pound, an amount equal to any pre-completion tax liability of the company that was not provided for in the completion accounts, together with the costs of dealing with it. That matters for three reasons.

• No valuation argument. The buyer does not have to prove the shares were worth less because of an unexpected PAYE assessment. It proves the assessment and the amount.
• Disclosure does not usually cut it down. A buyer is not taken to have accepted a known tax exposure simply because the seller mentioned it, which is the opposite of the position for tax warranties.
• A longer time limit, tied to HMRC’s own windows. For corporation tax, paragraph 46 of Schedule 18 to the Finance Act 1998 allows an assessment within four years of the end of the accounting period, extended to six years where the loss of tax was brought about carelessly and twenty years where it was brought about deliberately. Six or seven years from completion is therefore the normal negotiated period, and it is why tax claims are carved out of the shorter commercial warranty limit.

The tax warranties still earn their place alongside the covenant. They flush out information through the disclosure process, they cover matters the covenant does not reach such as the loss of a relief or an unexpected restriction on carried-forward losses, and they give a damages claim where the covenant’s trigger is not met. Buyers should have both and should not let a seller argue that one makes the other redundant.

Tax treatment of a payment under a tax covenant, and of the sale proceeds themselves, is a matter for your accountant or tax adviser. We set out the legal mechanics, and we work alongside your tax adviser on the drafting, but every seller should take their own tax advice on their specific position before signing.

What this means for your deal

If you are buying, the useful rule is that warranties protect you against the unknown and indemnities protect you against the known. Every issue your due diligence turns up should be met with a price reduction, a specific indemnity, or a condition that it is fixed before completion. Leaving it to the warranties means accepting a valuation argument you may never bring. This sits within the wider sequence we set out in our guide to buying a business in the UK.

If you are selling, your exposure is set long before the agreement is signed, by how thoroughly you prepare the disclosure letter and how firmly you hold the line on the cap, the thresholds and the notification period. Agreeing a specific indemnity is a decision to pay in full for that risk, so it should be priced, not conceded. Our guide to selling a business in the UK covers the process as a whole, and our wider note on corporate legal services explains how these pieces fit together across a company’s life.

Either way, the warranty and indemnity package is not boilerplate to be skimmed at the end of a negotiation. It is where the deal decides who pays if the business is not what it was said to be.

We advise buyers and sellers of owner-managed and SME businesses on share and asset sales, from heads of terms through to completion, and we are equally used to sitting on either side of the table. If you are negotiating a warranty schedule, weighing up an indemnity, or trying to work out whether the limitations you have been offered are reasonable, we can tell you quickly. Talk to our team about selling a business or buying a business, see our wider mergers and acquisitions and shareholders’ agreements work, or call us on +44 207 566 1188 or email info@gurvelegal.com for an initial conversation.

This article sets out the general legal position as at 6 October 2026 and is not legal advice on any particular transaction.