A group sells its foreign plant, the buyer takes over an asset insured for six years, and nobody raises insurance before signature because it looks incidental next to price, warranties and regulatory clearances. Three months later a measure occurs, the buyer notifies, and learns it is insured by nobody. This module deals with what happens then, and above all with what should have happened six months earlier in one meeting.
The governing principle is simple and rarely stated: an insurance policy does not pass with the property it covers. It binds an insurer to a named insured, over a described interest, and it was written in consideration of that particular person. The underwriter assessed this group's governance, its local partner, its conduct of earlier claims, the circle of its knowledge parties. A buyer is a different risk, of which it knows nothing, and nothing obliges it to accept that risk at the price it granted the seller.
The mechanical effect of the sale deserves precise description, because it creates a gap whose date is known in advance. On completion the seller ceases to have an insurable interest in the asset sold: its cover, on that asset, becomes purposeless, and it could no longer claim a loss it would not suffer. The buyer has no contract at all. Cover therefore does not degrade gradually, it stops at a precise hour, and that hour is the one at which everyone is celebrating the signing.
A distinction then governs real cases and it has already been met in relation to interposed holdings: who is the named insured? Where the policy is placed by the local company itself, selling that company's shares does not change the insured, which continues to exist and remains covered. Where it is placed by the parent, which is the commonest case in this field, selling the subsidiary leaves the policy in the hands of a seller with nothing left to insure. Two transactions alike in every respect produce opposite results, and the difference turns on a name on a cover page.
Even in the first case, a frequent clause reopens the question: change of control. Many policies provide for automatic termination, or require prior agreement, where control of the insured changes hands. The reason is the same as above, and it is legitimate: the insurer underwrote a shareholder as much as an asset. A buyer that checked the policy was in the local company's name and thought itself safe then discovers that the very same policy provided for its own extinction on the day of its arrival.
There remains a question sale agreements handle badly and which is often worth more than future cover: who owns the pre completion claims? A triggering event occurring before completion but notified after belongs in principle to whoever suffered the loss, that is the seller, unless the price already accounted for the loss, in which case the buyer will argue it is entitled to it. That allocation is settled in three lines in the sale agreement and almost never appears there, because both negotiating camps believe the other has dealt with it.
Finally one must say why the question cannot be recovered afterwards. Before signature the insurer has an interest in keeping an account: it can accept a novation, endorse a new insured, adapt the terms. After completion it has nothing left to keep, the buyer is an ordinary prospect, and it will offer a fresh policy at today's rate, on a country that may have deteriorated over six years. The gap between the two situations is not legal, it is one of bargaining position, and it flips the second the price is paid.
Three steps close this module and the first sets the timetable. Put insurance on the table at the letter of intent stage, alongside regulatory clearances, and not in the list of ancillary conditions. Obtain the insurer's written agreement, by endorsement or novation, as a condition precedent to completion rather than as a covenant to try. And write into the sale agreement who owns claims whose triggering event precedes completion, distinguishing those already notified from those not.
A European infrastructure fund acquires, on 30 April 2026, one hundred per cent of a local company operating a port terminal in a West African country, from a Spanish industrial group. The political risk policy, placed in 2020, names the Spanish group as insured and as insured object the interest held in the local company, for 175 million euros, expiring on 31 December 2027. The sale agreement mentions insurance in a general clause by which the seller undertakes to provide reasonable cooperation in transferring the policies. On 9 August 2026 a newly created port authority imposes on the terminal a capped tariff below operating cost and appoints an operations director. The fund notifies. What will it discover, and what remains?
It will discover that it is insured by nobody, and the mechanism is clean rather than unfair. The named insured is the Spanish group, not the local company, and the insured object is the interest that group held. Since 30 April the Spanish group no longer holds that interest: it has no insurable interest and would suffer no loss, so it cannot claim. The fund was never party to that contract. Cover did not degrade, it stopped on 30 April at the hour of completion, and the 9 August triggering event falls into the gap. The reasonable cooperation clause recovers nothing: it is an obligation of means over a transfer that did not happen, it does not bind the insurer, which is not party to it, and it creates no cover. One must then look at what remains, and three unequal things do. The first is to check whether the insurer was told of the transaction before completion and let it pass, which would open an argument about tacit acceptance; that is thin but it is verified in a day in the broker's correspondence, and if the broker wrote to the insurer in March it must be produced. The second is a claim against the seller on the cooperation clause, but it will be necessary to show the transfer would have been obtained had the seller behaved differently, a difficult proof since the insurer was free to refuse. The third is the most useful and it looks forward: ask the insurer at once for a fresh policy, knowing it will be dearer and will not cover the 9 August event, and above all check whether the fund holds other assets in identical situations. Finally the fund must be asked about price, because that is the real lesson: it paid for a terminal assuming cover it did not acquire, and that assumption had a value appearing nowhere in its model. The missing clause is one line long, the insurer's written agreement as a condition precedent to completion, and it was negotiable in April when the insurer still had an account to keep.
- 01A policy does not pass with the property: it binds an insurer to a named insured, written in consideration of that person, its governance and its handling of claims.
- 02Cover does not degrade, it stops at the hour of completion: the seller loses its insurable interest, the buyer has no contract.
- 03Everything turns on who the named insured is: a policy placed by the local company survives the sale of its shares, one placed by the parent stays with a seller that has nothing left to insure.
- 04A change of control clause can extinguish the policy on the very day the buyer arrives, including where the named insured does not change.
- 05The bargaining position flips the second the price is paid: before, the insurer has an account to keep; after, the buyer is an ordinary prospect.