The policy covers what was unknown. The formula is simple and means nothing until two questions are answered: unknown as of when, and unknown to whom. A deal has at least four dates that count, the signing of the confidentiality agreement, the signing of the sale agreement, the date the policy is underwritten and closing; and it brings together several circles of people, the deal team, the buyer's own directors, its advisers, and the target's managers who very often stay in place. Cover is defined by intersecting a date with a circle, and every wording draws that intersection its own way. It is on that intersection, and not on the size of the limits, that most declined claims turn.
The two dates that decide are signing and closing. Between them pass weeks or months, the time needed to obtain regulatory clearances. Warranties are given at signing and most often repeated at closing. The policy covers the warranties given at signing; cover for their repetition is a distinct question, and many wordings grant it only for matters ARISING during the interim period, not for matters DISCOVERED during it. The distinction looks byzantine and it opens a considerable gap: a fact that already existed at signing and is discovered before closing is no longer unknown on the day of repetition, and it is not new since it pre-existed. It therefore falls under neither.
The practical consequence is harsh and must be stated without softening, because it takes buyers by surprise at the worst possible moment. Whoever discovers a problem between signing and closing faces three doors, all of them bad. Say nothing and close, which may breach the duty of disclosure to the insurer and takes the item out anyway. Reopen the price, which the sale agreement almost never permits once signed. Or walk away, which a material adverse change clause rarely allows, those clauses being drafted to be all but unreachable. Insurance rescues none of those three doors, which is why the buyer's best position was to have found the fact before signing, while it could still be handled in the price.
The circle of people deserves the same attention as the dates, and it is negotiable. Knowledge is defined by a list of named individuals, often five to fifteen, together with a standard. The difference between actual knowledge and knowledge after reasonable inquiry is enormous, and it passes for a drafting nuance. The second standard imputes to the insured what it would have known had it asked, so that an individual who could have known is treated as knowing. Cutting the list to a few names and obtaining actual knowledge alone is one of the rare places where negotiation genuinely moves the contract's value, and it is a few hours' work done at the outset or never.
A difficulty peculiar to these deals concerns the target's managers who stay in place, which is the case in most acquisitions and almost always where they reinvest in the equity. Their pre-sale knowledge was the seller's; after closing, they run a company belonging to the buyer. Whether, and from when, their knowledge is imputed to the buyer receives different answers across wordings, and it decides whole files. A manager who knew of a problem, kept quiet, sold, stayed, and today runs the business on the buyer's behalf is a walking imputation problem, and generally appears on the list of individuals whose knowledge counts.
That leaves the most dangerous document in the file, and it runs to one page. At closing, the buyer signs a no-claims declaration, asserting that it has no knowledge of any breach of warranty. It is signed in haste, in the middle of fifty signature pages, often by someone who has not questioned the eight individuals on the list. Yet it converts the knowledge question into a warranty given by the buyer to its insurer: it is no longer for the insurer to prove the buyer knew, it is the buyer that asserted it did not. Signing that document while a team member carried an unresolved concern is the most common way of losing an otherwise well negotiated cover.
The conduct that avoids all of this requires neither talent nor money, only discipline, and it comes down to three steps. Keep a dated log of points discovered during the deal, with what was done about each, so that knowledge is an established fact rather than a contested reconstruction. Formally warn the named individuals that their knowledge is contractual, which most of them do not know, and give them a single instruction: anything touching a warranted subject goes up, without filtering and without a judgment about whether it matters. And circulate the no-claims declaration forty-eight hours before closing rather than at the signing table, so that somebody physically has time to put the question to those who might answer no.
A buyer signs the acquisition of a services group on March 12; closing is set for July 4, the time needed for competition clearance. The warranties are repeated at closing, and the policy covers that repetition only for matters arising during the interim period. Knowledge is defined as the actual knowledge of eight named individuals, among them the integration lead. On May 22, an operational director of the target mentions to that lead that a customer representing 14% of revenue has said orally that it will not renew its contract at expiry. The lead notes it in an internal memo, treats it as commercial noise, and does not escalate it. On July 4, closing takes place and the buyer signs the no-claims declaration. In September, the customer confirms its departure; the buyer claims 7 million under the warranty on customer contracts. What will it obtain?
Very probably nothing, and the date that loses the file is May 22, not September. The integration lead is on the list of eight: his actual knowledge is the buyer's, from the day he received the information. The fact pre-existed the March 12 signing, so it did not arise during the interim period, and the July 4 repetition does not cover it; it was no longer unknown at that date, and the original policy does not cover what the insured knew either. The point falls out on both sides at once. There is a further difficulty the buyer did not see as it signed: the July 4 no-claims declaration was inaccurate when signed, which is more serious than the declinature itself, since it can ground a challenge reaching beyond this single claim. The only available defense is to argue that on May 22 there was an informal remark and not an established fact; it is not absurd, it is the only case to run, and it will turn entirely on the drafting of the internal memo, which will be produced. What should have been done cost nothing: a single instruction to the eight named individuals, that anything touching a warranted subject goes up without filtering, and the no-claims declaration circulated forty-eight hours before closing rather than at the signing table. The point would then have surfaced in late May, while it was still possible to disclose it and argue about the price.
- 01Cover is defined by intersecting a date with a circle of people, and most declined claims turn on that intersection.
- 02A fact existing at signing and discovered before closing falls into a gap: it is no longer unknown and it did not arise during the interim period.
- 03Actual knowledge and knowledge after reasonable inquiry are not the same cover: the second imputes what would have been learned by asking.
- 04The target's managers who stay carry pre-sale knowledge inside a company that now belongs to the buyer.
- 05The no-claims declaration signed at closing reverses the burden: it is no longer for the insurer to prove the buyer knew.