The five previous lessons followed the asset: frozen by a designated partner, cut off from its market by Article 12f, held by the state that responds, stripped of its bank security, placed under temporary administration. This one moves to the other side of the table, to the underwriter who must decide whether to take such a risk and how to read it. It does not say what should be offered to an insured, which belongs to another profession; it says what a file exposed to sanctions must make visible before a decision is taken, and the three points on which most files stay silent.
The first point is the sanctions clause, and the London market has settled how to read it. On October 12, 2018, in Mamancochet Mining v Aegis ([2018] EWHC 2643 (Comm)), Mr Justice Teare examined a clause under which the insurer is not liable to pay to the extent that payment would expose it to a sanction. He held that an actual prohibition on payment is required, not a mere risk of sanction: the possibility of US secondary sanctions was not enough. He also held that the clause suspends the obligation to pay without extinguishing it. An underwriter writing this clause therefore knows what it does: it protects against prohibition, not against fear, and it lets the debt revive on the day the prohibition falls away.
What the clause does not settle is the interval. Between the moment a loss arises and the lifting of a measure, several years can pass, and the suspended debt sits on the insurer's balance sheet as a liability whose payment date is unknown. The London market has measured that interval in another line: the annual survey of credit and non-payment claims counts, for 2025, four claims not paid on time because of sanctions, 3 percent by number but 20 percent by value, the time needed to obtain a license. A suspended loss is therefore rare and heavy, and the reserve it calls for depends on how long the measure lasts, which nobody can predict.
The second point is the chain, which this course's lessons have shown from every angle. Who are the partners, and who controls them, under the three ownership and control tests? What goods does the asset receive or produce, and are they in the regulations' annexes? Through which banks do collections pass, and under which regimes are those banks? What security backs the flows, and what law binds the banks that confirm it? Name screening, which checks that the insured appears on no list, answers none of these questions. The no-re-export clause the insured already has its customers sign and the enhanced due diligence the packages impose are, in this work, sources of information as much as obligations.
The third point is accumulation, and it is what the losses of recent years revealed. The leased aircraft, the subsidiaries placed under temporary administration, the investments in a corridor country share one feature: a single act of state struck, on the same day, policies that nothing linked on paper. The insureds were different, the assets were different, and sometimes so were the lines, political risk, aviation war risks, trade credit. The key linking them was the state, and more precisely the relationship between that state and the insureds' side. A portfolio is therefore read by country, but also by pair of countries: what matters is less the host country's stability than the probability that an open conflict between it and the Union will trigger measures and retaliation.
One must add the case of secondary sanctions, which weighs on the underwriter itself. A European insurer covering a transaction lawful under EU law can find itself threatened by a foreign regime, most often the US one, that targets those who trade with its target. EU law does not require it to comply, and the blocking statute even prohibits compliance in some cases; but its access to the dollar, to its reinsurers and to its large clients depends on its conduct. That is why a market withdraws from a country its own law does not prohibit: the underwriting decision factors in a sanction that does not apply to the contract in law, but applies to the insurer in economic terms.
Three questions close this course, and they are the ones a file must make it possible to settle. What act, by which state, could deprive the insured of its asset, its market, its payments or its security, and would the policy see that act or exclude it as an act of the home country? What would become of a loss arising under a prohibition to pay, and for how long would the insurer carry a suspended debt? With which other risks in the portfolio does this file share the same key, that is, the same pair of states? A file that does not make it possible to answer these three questions is not incomplete on a detail: it does not yet say what is being insured.
A political risk underwriter reviews three submissions received the same week. The first concerns the Kazakh subsidiary of an Italian equipment maker, which imports computer numerically controlled machine tools from Europe. The second concerns a logistics terminal in Armenia, owned 60 percent by a French group and 40 percent by a local partner about whom the file says nothing. The third concerns a packaging plant in Turkey, wholly owned by a German group, half of whose output goes to customers in Central Asia. All three files contain name screening with no hits. The underwriter must say what is missing before any decision, without proposing anything to the insureds.
All three files passed name screening, and none yet makes it possible to answer this course's three questions. The Kazakh subsidiary depends on goods that Article 12f has already targeted for a neighboring country: what is missing is the share of its business that depends on goods in the annexes, and the policy's position on acts of the home country, since it is the Union, not Kazakhstan, that could cut that flow. The Armenian terminal has a 40 percent partner about whom nothing is known: what is missing is that partner's ownership and control chain under the three regimes, since a designation could tip the company if the partner controlled more than its stake shows, through an agreement or a right to appoint management. The Turkish plant has neither targeted goods nor an unknown partner, but it sells into Central Asia: what is missing is the banks its customers pay through and their position with regard to recent listings. One question remains that the three files raise together: they share the same key, the relationship between the Union and Russia, of which these three countries are trading neighbors. Taken separately, each looks modest; taken together, they form one exposure that the same sanctions package could trigger, and that is why they must be read side by side before being decided one by one.
- 01Since Mamancochet (2018), the 'would expose to a sanction' clause requires an actual prohibition on payment, not a mere risk, and it suspends without extinguishing.
- 02A debt suspended by sanctions is rare and heavy: in London in 2025, claims delayed by sanctions made up 3 percent of credit and non-payment claims by number, but 20 percent by value.
- 03Name screening says nothing about the chain: partners and control, targeted goods, banks, security and the law that binds them.
- 04The accumulation key of a portfolio exposed to sanctions is a pair of states, not a country or an insured.
- 05Secondary sanctions apply to the insurer in economic terms even when they do not apply to the contract in law.