The same political risk can be carried by a private insurer or by a public export credit agency, and that choice is ordinarily presented as a question of price and capacity. That is the wrong axis. A private insurer underwrites to earn a margin; a public agency underwrites because its state has decided to support its exports, and insurance there is an instrument of economic policy. This difference in nature does not stay at the level of intentions: it decides who can be insured, at what price, and above all what happens once the indemnity has been paid, which is the subject least discussed when the risk is being placed.
The first consequence is that eligibility is conditioned on the national interest, and principally on a share of national content in the contract. That is a condition of access, and it is also a trap whose timing must be understood. The share is measured on the contract as presented at underwriting, but an industrial contract lives: a subassembly is bought elsewhere for cost reasons, a subcontractor changes, a plant is relocated. The national content share then falls without anyone noticing, because it is taken in a department with no reason to suspect that a cover depends on it. The eligibility condition is a continuing obligation rather than a snapshot, and it is verified at the time of the loss, that is at the moment when nothing can be corrected any more.
Price is not free either. Public agencies are bound by international rules setting minimum premium rates and maximum repayment terms, whose purpose is to stop states subsidizing their exports through cheap insurance. The result is a situation exporters understand poorly: the public rate is a regulatory floor rather than a market price, the private market can therefore be cheaper on some risks, and the agency may find itself unable to offer the tenor the buyer wants, not from lack of appetite but from applying a rule that does not concern it alone.
It must also be known that the boundary between the two markets is drawn by regulation and that it moves. The notion of marketable risk reserves short-term risks on certain countries to the private market, and bars public agencies from covering them, on the basis that private supply is deemed to exist. When a crisis withdraws private capacity, that classification can be suspended and public cover reopens. An exporter's access to cover therefore depends on an administrative classification that changes with market conditions, generally in the right direction but with a lag, and which closes again afterward. An exporter can thus lose or regain access to cover without its own risk having moved at all.
Then comes the difference that matters most and is discovered latest: the two systems do not recover in the same way. When a private insurer has paid on a sovereign default, it recovers like any private creditor, by negotiation or arbitration, with the immunity obstacles already described. When a public agency has paid, the claim becomes that of one state against another. It enters the intergovernmental treatment of sovereign debt, where creditor states collectively reschedule a debtor state's obligations, requiring other creditors to accept comparable treatment. The exporter's claim is therefore absorbed into a sovereign debt negotiation to which it is not a party, on a timetable set by diplomacy.
That absorption cuts both ways, and this is what must be weighed before choosing. The public route holds leverage no private insurer will ever have: a state seeking new financing, new export guarantees or a treatment of its debt must deal with its creditor states, which is worth far more than an unenforceable award. But it is slow, collective, and the exporter has no seat: its recovery arrives as a share of a rescheduling, spread over years, decided without it. The choice between public and private is therefore a choice between two recovery systems, one with leverage and no control, the other with control and no leverage. This is almost never discussed at placement, where the conversation is about premium and insured percentage.
That leaves the frequent case of a mixed structure, where the exporter takes public cover for political causes and private cover for commercial ones. The point to see is not only that the two contracts define political cause differently, but that recovery splits. One and the same unpaid receivable may require notification to both insurers, and if both pay, each is subrogated for its share into a claim that will follow two distinct routes, one intergovernmental and slow, the other private and faster. Any later settlement with the debtor state will therefore have to satisfy two subrogated parties whose interests and horizons differ, which makes a global agreement appreciably harder to reach than with a single insurer.
A French industrial equipment maker signed a 30 million euro contract with a foreign public buyer. It obtained cover from its country's public agency on the basis of a 68% national content share, the policy conditioning cover on maintaining national content of at least 60%. During performance, the procurement department moved manufacture of a subassembly to an Asian supplier, saving 400,000 euros; the national share fell to 51%, and nobody told the insurer. In January 2026, the buyer stops paying on instruction from its supervising ministry, a political cause being established. The unpaid balance is 12 million. The agency asks for the national content position at delivery. What happens?
The file is in danger for a reason having nothing to do with the insured risk, and that is what makes it instructive. Nobody committed an isolated fault, and that is precisely the problem: national content was treated as an admission formality, settled once and for all at underwriting, when the policy makes it a continuing condition verified at the time of loss. Procurement took a commercially rational decision, saved 400,000 euros, and imperilled 12 million of cover, without either department involved having any means of connecting the two. Two things remain to be examined, and the order matters. First the wording: is the share measured at signature, at delivery, or on average across the contract, since those three readings do not give the same result and the first would save the file. Second, the duty to declare material changes, almost always present: a decrease declared at the time could have been accepted, possibly against an adjustment, whereas an undeclared decrease is analyzed quite differently. And even if eligibility survives, the maker must be told what it does not know: being indemnified by a public agency will not make it recover any faster, because the claim will pass into the intergovernmental treatment of debt, on a timetable of several years that must be provisioned for now.
- 01A public agency insures to support exports, a private insurer to earn a margin, and that difference decides eligibility, price and recovery.
- 02National content is a continuing condition verified at the time of loss, not a formality settled at underwriting: a sourcing change can cost the cover.
- 03The public rate is a regulatory floor rather than a market price, and the boundary between the two markets moves with an administrative classification.
- 04The public agency recovers through intergovernmental debt treatment: much leverage, no seat for the exporter, and a timetable in years.
- 05In a mixed structure, two subrogated insurers follow two different recovery routes, which makes any later settlement harder to conclude.