An investor who sees the measure coming sells before it lands. It transfers its interest to a local buyer at a third of its value, and saves a third rather than nothing. Six months later the decree appears, and it notifies the difference to its insurer. The loss is real, it can be quantified, and it is not covered. This module explains why, then says what would have had to be established before signing for it to become covered, because that is the only part that is any use.
The reason for the refusal is not an exclusion, it is deeper and it lies in what a policy covers. A political risk cover indemnifies a deprivation suffered, not a price accepted. A sale is a voluntary act by which the seller receives consideration it judged preferable to keeping the property; it was not deprived of its asset, it exchanged it. No wording turns that act into a loss, and this is not market harshness but the very definition of what is insured.
To this must be added a consideration the market does not always say aloud and which settles its position. If a discounted exit were recoverable, every investor in a deteriorating country would have an interest in selling cheaply to a related party and claiming the difference. The loss would become a choice, its amount would be set by the party suffering it, and the cover would fund a transaction rather than indemnify a loss. Moral hazard there is total, which is why the door is shut by construction and not by appetite.
The useful question remains, the one the brief poses: what had to be established before selling? Three things, and they hold together. The first is that a triggering event had already occurred. A creeping expropriation under way at the time of sale changes the nature of the transaction entirely: control was already lost, the sale is no longer a commercial choice but the realisation of a loss already constituted, and the price gap measures that loss rather than creating it. Everything therefore turns on the date of the triggering event, as almost everything does in this field.
The second is that the sale was a mitigation measure. An insured limiting its damage performs an obligation its policy imposes, and it must be able to show it in the proper form: prior notification to the insurer, a statement of the alternative, costing of both branches, and if possible the insurer's written position. That process looks heavy to an executive in a hurry to leave; it takes a week and it is the difference between a sale endured and a sale chosen. An insurer often prefers to cap its loss through a partial exit rather than face a total expropriation, and it can say so only if asked.
The third is that the price was the best obtainable. A sale to a single local buyer, with no competition, is very hard to defend: it is indistinguishable, seen from outside, from a taking dressed as a sale, and all the more so if the buyer turns out to be close to the administration. Even a short process, a few buyers approached, written offers kept, turns a price endured into a price established. That is not file formality, it is what allows one to say the discount comes from the country's situation and not from the buyer's identity.
The tension all this produces must be named, because an honest adviser does not hide it. An investor that does not sell and is expropriated has acted well as regards its policy and may have acted badly as regards its business: it will have waited two years for an indemnity that a third of the price gave it at once. An investor that sells has done the reverse. The policy does not reward commercial prudence, it indemnifies a deprivation, and it is better to say so plainly to an executive than to let it believe it can have both.
Three steps close this module and all are taken before signature. Establish and date what has already happened, because a prior triggering event changes the nature of the sale. Notify the insurer before selling, setting out the costed alternative and asking for its position in writing. And run a competitive process, even brief, even fruitless, because an established price can be defended and an accepted price cannot.
A European group held 55 per cent of a fuel distribution company in a West African country, valued at 180 million euros in summer 2024. From September 2024 the regulator imposes a capped sale price below supply cost, refuses to approve two successive chief executives and opens three tax audits in five months. In March 2025 the group transfers its interest to a local investor for 61 million euros, with no competitive process, the buyer having been introduced by an adviser to the ministry. In September 2025 a decree nationalises the whole sector with compensation set at net book value. The group notifies a 119 million loss to its insurer, which declines on the ground that it sold. What is its position worth?
The insurer's refusal is right on the principle it invokes and it is not right on this file, and the difference must be established rather than contested wholesale. On the principle, a policy does not indemnify a price accepted: had the group sold in September 2024 when nothing had happened, there would be nothing to say. But that is not what occurred. Between September 2024 and March 2025 the group suffered a sale price capped below supply cost, two refusals to approve a chief executive and three tax audits. That is the classic sequence of a creeping expropriation, and the price capped below cost is the strongest link because it makes the operation structurally loss making without any title having been touched. If effective control was already lost by March 2025, the sale is no longer a commercial choice but the realisation of a loss already constituted, and the price gap measures it instead of creating it. That is where the file is won or lost, and it is shown by a dated chronology, board minutes, approval requests and audit notices. Two serious weaknesses must be addressed and not sidestepped. The group did not notify its insurer before selling, although the facts from September to March were notifiable and the sale could have been presented as a mitigation measure; that silence will have to be explained, and the ordinary explanation, that one did not wish to sour the discussion, is human and weak. The absence of a competitive process is heavier still: a sale to a single buyer introduced by a ministry adviser, at a third of value, is very hard to distinguish from a taking dressed as a sale, and the insurer will certainly say so. One must therefore establish what was attempted, even informally, and if nothing was, concede it and put the whole effort on the date of the triggering event, which is the only solid ground. One point finally deserves flagging to the group: nationalisation compensation set at net book value is far below 180 million, meaning that by not selling it would probably have obtained less than 61 million from the state, and that its commercial decision was likely sound. It was sound and it was badly prepared, and that is exactly what this file costs.
- 01A policy indemnifies a deprivation suffered, not a price accepted: a sale is a voluntary exchange, and no wording turns it into a loss.
- 02If a discounted exit were covered, the loss would become a choice and its amount would be set by the party suffering it: the door is shut by construction, not by appetite.
- 03What changes everything is the date of the triggering event: an expropriation already constituted makes the sale the realisation of a loss, and the price gap measures it instead of creating it.
- 04A sale presented as mitigation presupposes PRIOR notification, the costed alternative and the insurer's written position: that takes a week.
- 05A price established by competition can be defended, a price accepted from a single local buyer cannot, whatever the reality of the constraint.