The commercial pitch for a warranty and indemnity policy says it covers breaches of the seller's warranties. That is true, and it is not the sentence that decides a claim. What decides is the list of what falls outside. An adviser who approaches that list as a set of negotiating points has the method wrong and wastes the client's time at the worst moment, because a large part of those exclusions is structural: it follows from what insurance is, not from an underwriter's appetite. Knowing how to sort structural exclusions from negotiable ones is probably the most useful thing an adviser brings to this subject.
The first structural family is the known. Insurance covers uncertainty; a fact placed in the data room, flagged in a due diligence report or known to the deal team is no longer uncertain for the buyer, and it falls out. This is not an underwriter being difficult: covering the known would make the policy a price adjustment mechanism, and price is precisely what the parties negotiate between themselves. The consequence unsettles buyers the first time they meet it. The better the review, the more it finds, and the more falls outside cover. Discovering a problem moves it from the domain of insurance into that of commercial negotiation.
From this comes the paradox that governs the whole product, and that must be understood before any review scope is set. A thin review leaves subjects unknown and therefore theoretically insurable, but an underwriter will not cover what nobody examined either. A thorough review discovers problems, which then fall out as known. Cover is therefore greatest in a narrow band: subjects that were competently examined and where nothing adverse was found. Put another way, and this is the sentence to retain from this module, the policy insures the risk that a good review was wrong. It insures neither the risk of not having looked, nor the risk of what was seen.
The second structural family is the forward-looking. Price adjustments, working capital true-ups, earn-outs, projections and forward-looking statements are excluded. The reason is not their size but their nature: they are not assertions about a state of facts that could be true or false at signing, they are allocations of a future between the parties. Insuring them would hand an underwriter the arbitration of a commercial negotiation, and the loss would be indistinguishable from the buyer having paid too much. The same logic explains why many wordings restrict indemnification of lost profits computed on a multiple, which would turn a shortfall in earnings into an insured loss through the sole effect of a valuation formula.
The third family is of another nature, and this is where an adviser genuinely earns a fee. Environmental liability, transfer pricing, pension underfunding, and certain sanctions and bribery matters are frequently excluded depending on the market and the target. Those subjects are not uninsurable: they are excluded because assessing them requires specialist review that a deal timetable rarely allows. They therefore move. With a specialist report and enough time, they come back inside the perimeter, sometimes with a sub-limit or in a separate policy. The right question in front of an exclusion is never whether it can be deleted, it is which family it belongs to, because the answer dictates whether to negotiate, to commission a report, or to say nothing and provision.
It must then be known that the exclusions are not all in the exclusions article, and this is the point quick readings systematically miss. The definition of Loss, the definition of Warranty, the de minimis threshold and the retention each remove exposure without ever presenting themselves as exclusions. A wording that defines Loss as what would be recoverable at law from the seller under the sale agreement silently imports every limitation of that agreement: its caps, its time bars, its non-accumulation clauses. The buyer believes it holds a twenty-four million policy and in fact holds a claim bounded by a contract it negotiated downward itself in exchange for the insurance. Reading the exclusions article alone gives a false picture of the cover, and that false picture is the one presented to investment committees.
That leaves the timing consequence, which explains why this module comes before the one on due diligence. Exclusions are settled at the end of underwriting, under time pressure, when the deal must sign and nobody has any margin left. An adviser who discovers them in the final forty-eight hours has no leverage; one who maps them at the outset against the review workplan can steer the work toward the subjects whose exclusion would cost the most. A policy's exclusions are thus decided, in practice, by a review scope settled weeks earlier by people who did not know they were writing them.
A fund acquires an industrial group at an enterprise value of 240 million euros. The warranty and indemnity policy carries a 24 million limit and a 1.2 million retention. With a six-week deal timetable, the due diligence workplan covers legal, financial and tax, and for environmental matters is limited to a desktop review with no site visit and no sampling. The data room contains, in an annex to a 2019 report, a soil study noting hydrocarbon traces at an owned site; nobody summarized it in the headline reports. The policy excludes environmental matters, for want of specialist review, and separately excludes facts disclosed in the data room. Fourteen months after closing, the authorities impose remediation costed at 9 million. The buyer notifies a claim. What can it expect?
Nothing, and what makes this file instructive is that there are two independent grounds: removing one would not have sufficed. Environmental matters are excluded because only a desktop review was done, and the underwriter did not decline the risk, it declined a subject nobody had examined. Even if that exclusion fell away, the 2019 soil study sat in the data room: the fact was disclosed, therefore known, therefore outside cover under the first family of exclusions. The temptation is to argue that nobody read the annex and that the buyer did not actually know. That is the wrong battle and it is almost always lost, because these policies assess disclosure by reference to what was fairly made available in the data room and not to what the deal team actually read: an underwriter does not insure the buyer's reading habits, and the opposite rule would make disclosure unverifiable. The decision that costs 9 million was taken six weeks before signing, when the review scope was settled. On an industrial target owning its sites, at an enterprise value of 240 million, environmental liability is a subject whose exclusion costs more than the review that would have avoided it. The choice of a desktop review was made to hold a timetable, by people unaware they were drafting an exclusion. The workplan wrote the policy, not the other way round.
- 01What decides a claim is not what the policy covers but what falls outside it, and much of that list is structural rather than negotiable.
- 02The known is excluded because insurance covers uncertainty: discovering a problem moves it from cover into price negotiation.
- 03Cover is greatest in a narrow band, subjects competently examined where nothing was found: the policy insures the risk that a good review was wrong.
- 04The forward-looking falls out on principle, adjustments, earn-outs and projections, because it allocates a future instead of recording a fact.
- 05The exclusions are not all in the exclusions article: the definition of Loss silently imports the caps and time bars of the sale agreement.