Classifying the fact: four covers that overlap

A political risk policy does not cover 'political risk'. It covers four or five named facts, each with its own definition, triggering event and procedure. A claim therefore always starts with a question of classification, and that question decides what follows far more reliably than the amount at stake. The four usual facts are expropriation, currency non-transfer, political violence, and breach of contract by a sovereign entity. They look distinct on a brochure. In a real file, they almost always overlap.

Expropriation is the oldest and best mapped. It covers the act by which a state deprives the insured of its asset or of control over it, whether by nationalization, requisition, or through a series of measures reaching the same result without ever formally transferring ownership. That second form has a name, creeping expropriation, and it is the one that creates every practical problem. A plant that successively loses its export license, its access to foreign currency, then its right to appoint its own director has been expropriated by no decree at all: it was expropriated by accumulation.

Non-transfer covers something else, and the confusion is common. It is not about the value of the asset but about the movement of cash: the insured holds funds, in local currency, and can neither convert them nor take them out of the country. The central bank has confiscated nothing. It has closed the window. The distinction is clean in theory and blurred in practice, because cash blocked long enough loses its value to local inflation, and the insured then ends up dispossessed without any expropriation having occurred.

Political violence raises no classification problem but a boundary one. A riot that destroys a warehouse falls under that cover; the same riot that pushes the state to impose a six-month curfew, which ruins the operation, destroys nothing at all. The damage is economic rather than physical, and policies handle those two situations in different sections, sometimes in different contracts.

Sovereign contract breach is the most recent and the most contested. It covers the case where the buyer, the grantor or the guarantor is a public entity that stops performing. The difficulty is that any unpaid invoice from a state-owned company looks like a commercial default: same invoice, same delay, same silence. What moves the file from trade credit to political risk is not the nature of the debtor but the CAUSE of the default. A badly run state-owned company that does not pay is a credit risk. The same company, forbidden by its supervising ministry from paying a foreign supplier, is a political risk.

When several classifications overlap, two practical rules dominate. The first is that policies almost always carry a non-accumulation clause: one triggering event opens one indemnity, and the most specific cover applies, not the most favorable. The second is that the classification retained drives the date of the triggering event, therefore the start of the waiting period, therefore the year the claim attaches to. A file classified as non-transfer and a file classified as creeping expropriation on identical facts do not settle in the same year.

This is why classification is not an administrative formality to be settled once the loss has been quantified. It is worked first, documented while the facts are unfolding, and defended. An insured who discovers at notification that it kept no dated record of the central bank's successive refusals has lost the one piece of evidence that would have fixed a triggering event.