A warranty and indemnity policy sits between two documents drafted under different laws, above a group spread across several countries. Three distinct questions of law arise, and the habit is to collapse them into one: which law governs the sale agreement, which governs the policy, and which governs the place where the loss occurs. Those three answers have no reason to coincide, and each governs something different. Disentangling them is the first task in a cross-border deal, and it is ten minutes of work that saves months.
The dividing rule is as follows and it is easy to retain. The warranties come from the sale agreement and their meaning is construed under that agreement's law: whether a statement about the accounts was false is a question of sale agreement law. The insured's obligations toward its insurer, notification, disclosure, conduct of the file, belong to the law of the policy. A breach is therefore assessed under one law and compliance with one's own duties under another. That split is normal and it commands a discipline in pleading: the meaning of a warranty is a sale agreement question, the admissibility of a claim is a policy question, and mixing the two in one submission is the surest way to muddle a strong file.
It is on the disclosure standard that legal traditions collide most visibly, and the stake is quantifiable. Some systems treat information as disclosed only if it was fairly disclosed, with enough detail for a reasonable reader to identify the problem. Others accept that a general deposit in the data room discloses everything it contains. The policy imports whichever standard the sale agreement adopted, without arguing about it. The same data room, with the same documents, therefore covers or does not cover according to a clause negotiated by lawyers thinking about risk allocation between the parties rather than about the extent of an insurance cover.
A group operating in several countries then raises a scope problem the underwriter settles bluntly. It requires that local counsel examined each material jurisdiction, and where none did, that jurisdiction falls out. Covered countries are enumerated in the policy; they are not assumed. A group operating in eleven countries whose review covered four is insured in four, and the rest is not covered at a reduced rate, it is not covered at all. That enumeration is one of the rare places where cover can be seen at a glance, and it is nonetheless rarely read by those who decided the review scope.
Then comes the question that costs the most and that almost nobody asks before a claim: the tax treatment of the indemnity. An insurance payment received by the buyer may be taxable in its hands, while the loss it compensates sits in a foreign subsidiary and is deductible there, or nowhere. An indemnity of ten million can thus be worth seven million net to repair a ten million loss. Two mechanisms answer this. A gross-up clause, by which the insurer pays an amount such that the insured recovers the sum after tax, which insurers accept reluctantly and price. Or treating the payment as a reduction of the purchase price, more efficient where the applicable law permits it, but requiring the seller's cooperation.
That second route carries a timing contradiction that must be seen clearly, because it is the heart of this module. The most tax-efficient solution requires an agreement with the seller, and it must be written into the sale agreement, that is before anyone knows there will be a claim. Yet once the price is paid and its liability extinguished, the seller has no interest left in signing anything, nor any obligation to. The mechanism most useful to the buyer must therefore be negotiated at the moment it appears least necessary, with a party that will grant it readily then and never afterward. It is exactly the kind of clause struck from a draft to save time on a point nobody cares about on signing day.
Two technical points close the subject. The policy currency and that of the loss often differ, and the conversion date is settled in the contract or one replays a discussion other lines know well. And above all, the entity suffering the loss is not always the insured: if the buyer is a holding company and the loss strikes an operating subsidiary, the buyer's loss is the diminution in value of its shares, not the subsidiary's cash outflow. Those two figures are not equal and the second is often smaller. Good wordings say so expressly; the others leave the question open until the day it is worth several million. As for what actually gets negotiated in practice, the ranking is short and honest: the disclosure standard, the list of jurisdictions, and the tax gross-up. The law of the policy and the forum rarely move, and an adviser who spends the negotiation on them has optimized the wrong clause.
A French buyer acquires a group at an enterprise value of 260 million euros, with operating subsidiaries in Germany, Poland, Brazil and Vietnam. Legal due diligence was run by local counsel in France, Germany and Poland; for Brazil and Vietnam, only a summary note from the target group's own legal department was provided. The policy enumerates the covered jurisdictions and excludes Brazil and Vietnam. The sale agreement is governed by English law, with a fair disclosure standard. Thirteen months after completion, two facts emerge: the Vietnamese subsidiary has operated without a required environmental permit since 2019, at a compliance and penalty cost of 4.1 million; and the German subsidiary carries a 6.3 million pension underfunding, of which the data room contained only one line in the annex to an actuarial report. What does the buyer recover?
Nothing for Vietnam, and probably the 6.3 million for Germany. Vietnam is an excluded jurisdiction, and the exclusion follows directly from the absence of local counsel: a note from the target group's legal department is not a review report, it is not capable of being relied on by anyone and gives the insurer no recourse. The tempting argument, that the group knew its own operations better than any firm could, turns around exactly: internal knowledge is knowledge, which excludes, and not due diligence, which would have covered. It is the same mechanism as reliance, seen from another side. Germany turns entirely on the disclosure standard. One line in the annex to an actuarial report probably does not fairly disclose a 6.3 million underfunding, for want of detail sufficient for a reasonable reader to identify the problem: under that standard the disclosure fails and the claim stands. One must measure what that means: had the sale agreement provided that a deposit in the data room amounts to general disclosure, the same documents would have destroyed the claim. A single clause is worth 6.3 million here, and it was negotiated by lawyers allocating risk between the parties without thinking about the cover. That leaves the point nobody addressed and which will cost: the indemnity will be paid to the French holding company, where it may be taxable, while the pension liability is German. With no gross-up clause and no treatment as a price reduction, the buyer will net appreciably less than 6.3 million, and it is too late to agree it with a seller already paid and released.
- 01Three laws govern the matter: the sale agreement's for the meaning of the warranties, the policy's for the admissibility of the claim, the local one for the loss.
- 02The policy imports the sale agreement's disclosure standard: the same data room covers or does not cover according to a clause negotiated elsewhere.
- 03Covered jurisdictions are enumerated, not assumed: a group operating in eleven countries and reviewed in four is insured in four, not at a reduced rate.
- 04An internal note is not due diligence: it manufactures knowledge, which excludes, without manufacturing reliance, which would have covered.
- 05The tax treatment of the indemnity is settled before a claim exists, with a seller that will grant it readily then and will have no interest in doing so afterward.