An exporter discovering credit insurance almost always wants the same thing: to cover the two or three buyers that keep it awake and leave the rest. It is the most natural request in the world and it is the most expensive per insured name, and the reason is not commercial. A whole-turnover policy, taking the entire book, and a single-buyer policy do not buy the same thing: the first buys pooling, the second buys a chosen exposure. The insurer knows this, it prices it, and the price gap the exporter takes for greed is in fact the measure of what it has itself removed from the contract.
The mechanism is called adverse selection and there is nothing moral about it. If the insured chooses which buyers it insures, it chooses the riskiest, not out of malice but because it knows its book better than the insurer does and it would be unreasonable to do the opposite. The loss experience of a set sorted that way is therefore structurally worse than that of the book it came from. A whole-turnover policy removes the sorting by taking everything: good buyers there finance bad ones, the insurer gets a representative population, and the cost per buyer falls. That is pooling at its simplest, and it is why whole-turnover policies exist.
The second difference lies in what the insurer can do on its side, and it benefits the insured more than it realizes. On a whole-turnover policy, the insurer follows an entire population, cross-checks its information with that of its other insureds, and sees incidents arriving on buyers the exporter has not yet identified as fragile. That monitoring service has no equivalent on a single-buyer policy, where the insurer looks at one name only. Many exporters thus buy a cover and believe they are buying only a reimbursement, when the information received during the year is often worth more than the indemnity, because it prevents the shipment that will not be paid for.
One must nevertheless resist concluding that the whole-turnover policy is always the right answer, because it carries a cost the exporter measures badly. It requires declaring the entire turnover, including on buyers one would never have insured, and therefore paying on risks one considers nil. It imposes a management discipline, periodic declarations and observance of the limits granted on each buyer, on pain of forfeiture. For an exporter of which ten percent of turnover goes to one difficult country and the rest to safe domestic markets, it can amount to insuring on a large scale what does not need it.
Between the two, the market offers intermediate forms that must be known so as not to reason in all-or-nothing terms. A policy can be limited to a geographical zone, to a product line, or to buyers above a given outstanding, which reconstitutes a population without requiring the whole. A self-insurance floor on small buyers produces the same effect by dispensing with declaring a multitude of minor receivables. These carve-outs are accepted where they rest on an objective and stable criterion, and refused where they leave the insured a buyer-by-buyer choice, which brings adverse selection back through the window.
The single-buyer policy keeps uses where it is the only solution, and they must be recognized. An exceptional contract, far larger than usual turnover, is not handled inside a whole-turnover policy calibrated on regular flows. A one-off transaction to a country the whole-turnover policy excludes calls for dedicated cover. A receivable intended to be assigned to a bank often requires a named cover the lender can read. In those cases, the extra cost per name is not a penalty: it is the price of an exposure nothing dilutes.
The method of choice therefore comes down to three questions asked in this order, and none of them is about rates. The first is representativeness: does what I propose to insure resemble my book, or is it its peak? The second is service: do I need continuous monitoring of my buyers, or only reimbursement on a name I know? The third is the receivable's destination: does it stay with me, or must it be readable by a third party? An exporter answering those three questions knows which of the two structures suits it before having asked for any quotation at all.
An exporter turns over 40 million euros. Three customers account for 14 million, one of them, in a country under strain, for 6. The other forty, spread across the European Union, account for 26 million and have produced no incident in eight years. The finance department asks for a policy on the three large customers only. The broker comes back with two proposals and comments only on the cost gap, appreciably unfavourable to the solution requested. General management, finding the gap unjustified, is considering insuring nothing at all. How should the question be framed?
The cost gap is not greed, and reading it that way leads to the worst of the three possible decisions. What the finance department proposes to insure is not a sample of its book, it is its peak: three customers chosen because they cause worry, one of them in a country under strain. A set sorted that way has structurally worse loss experience than the book it came from, and the insurer is merely observing it. The question to ask is therefore not why this solution costs more, but whether the sorting is right. It probably is not, for a reason the price debate conceals: the forty European customers with no incident in eight years are not what one insures, they are what makes the other three insurable at a reasonable cost, and excluding them removes from the contract exactly what was holding it up. Two routes remain open and only one is bad. Carving out on an objective and stable criterion, by zone or by minimum outstanding, reconstitutes a population without requiring the whole and can be negotiated; leaving the choice buyer by buyer brings adverse selection back and will be refused or priced as such. Isolating the 6 million customer in dedicated cover is defensible where the exposure is one that nothing dilutes, particularly if the receivable must be readable by a bank. Insuring nothing, by contrast, amounts to keeping a 6 million exposure in a country under strain because a rate gap was misread.
- 01A whole-turnover policy buys pooling, a single-buyer policy buys a chosen exposure: they are not two versions of one product.
- 02Adverse selection is not a moral matter: an insured choosing its buyers chooses the riskiest, because it knows its book.
- 03Continuous monitoring of a whole population is often worth more than the indemnity, because it prevents the shipment that will not be paid for.
- 04A carve-out on an objective, stable criterion can be negotiated; a buyer-by-buyer choice brings adverse selection back and will be priced as such.
- 05The single-buyer policy remains the only solution for an exceptional contract, an excluded country, or a receivable a lender must be able to read.