Commodity trading is financed on a logic that is not that of corporate credit and must be grasped before applying cover to it. A lender financing a cargo of copper or grain does not look first at its borrower's balance sheet: it looks at the goods, at the sale contract with a solvent final buyer, and at the mechanism by which the proceeds of that sale will repay it before passing through the borrower's accounts. This is called financing a transaction rather than a company, and the structure is reputed safe because it unwinds by itself. Political risk is precisely what stops it unwinding by itself.
Three structures coexist and do not carry the same risk. Export prepayment advances funds to a producer before the goods even exist, against an undertaking to deliver; it is the most exposed, since it finances future production inside the risk country. Inventory financing backed by warehouse receipts bears on goods that exist but are physically located at the debtor's. Financing goods in transit is best placed, since the property has left the territory. These three positions are distinguished by a single question: where are the goods the day the country closes.
The central mechanism of these structures is the offshore escrow account into which the final buyer pays directly, and from which the lender repays itself before the balance flows up to the seller. It is that account that provides the security, more than pledges over the goods, and it is what political risk attacks. Exchange regulations requiring repatriation of export proceeds, an obligation to surrender part of the currency to the central bank, or an instruction to buyers to pay locally, hollow out the mechanism without touching the contract. The structure therefore rests on a rule of the seller's country, which can change by circular.
A second vulnerability is added, peculiar to this business: the goods themselves can be attached, requisitioned and sometimes barred from export. A state short of a product on its domestic market bans its departure, and that is a frequent measure on food and energy. The lender then holds a pledge over goods that exist, which it can neither move out nor sell to the final buyer, and whose local value bears no relation to the contract price. The loss is real and resembles no ordinary non-payment: there is no refusal to pay, there are goods immobilized by a public decision.
Cover for these transactions therefore combines guarantees ordinarily presented separately, and it is the assembly that counts. Non-transfer answers the blocking of proceeds, confiscation or requisition answers seizure of the stock, an export ban falls under contract frustration or government action cover depending on wordings, and the producer's default remains ordinary credit risk. A transaction believed to be covered can thus turn out bare on the very mechanism that operated, because the policy was bought on the family of risk rather than on the concrete unwinding of the deal. The useful reading starts from the structure and checks each of its breaking points.
One methodological point deserves saying because it distinguishes these files: here, proof of loss is largely physical. Commodity prices are public and dated, cargoes are traced by bills of lading and inspection reports, stocks are controlled by third-party holders, and the public decisions blocking them are written. A well-kept file is therefore easier to establish than on an ordinary commercial receivable, provided the documentary chain has been preserved, which means independent inspections rather than the debtor's certificates.
What is retained is an inversion of the gaze. On these transactions, one does not ask whether the debtor is solvent, one asks where the goods are, through which account the money passes, and what public decision can interrupt either. Three checks follow: the location of the settlement account and the exchange regulations applying to it, the physical location of the goods at each stage and the possibility of an export block, and the consistency between the guarantees bought and those two breaking points. A well-built transaction withstands a default; what it must withstand is a circular.
A European bank prepays 11 million euros to an East African agricultural producer, repayable from the proceeds of a harvest sold forward to a European trader. The contract provides for the trader to pay into an escrow account in London, from which the bank repays itself before the balance flows up to the producer. The harvest is good and stored in a warehouse under a third-party holder's control. On August 12, three weeks before loading, the government bans export of this commodity for six months to supply the domestic market, and requires existing stocks to be sold locally at an administered price representing 40 percent of the world price. Where is the loss?
One must first observe that none of the security mechanisms failed and that the bank loses anyway. The producer has not defaulted, the harvest exists, the third-party holder controls it, the trader is solvent, and the London escrow account is intact: it will simply never receive anything, since the goods will not leave. That is the hallmark of these structures, whose real security is automatic unwinding rather than pledges: they withstand a default and not a circular. The loss therefore sits in two distinct places that must be characterized separately, because they do not fall under the same cover. The export ban is a public measure preventing performance of the sale contract, and it is handled as contract frustration or government action depending on the wording bought, not as non-transfer, since no proceeds are blocked on the way out. The forced sale at an administered price is another thing: it is a deprivation of value on the pledged goods, falling under requisition or a measure equivalent to expropriation, and it is quantified on the gap between the dated world price and the 40 percent received. The bank's position therefore depends entirely on what it bought, and cover purchased on the non-transfer family would leave it bare on both mechanisms that operated. What works in its favour is the proof: public dated prices, the third-party holder's inspection reports, a written decree. The file is solid on the facts; it is fragile on the policy's perimeter.
- 01These structures finance a transaction and not a company: their real security is automatic unwinding, not pledges over the goods.
- 02Three positions are distinguished by one question: where are the goods the day the country closes, at the producer, in warehouse or in transit.
- 03The offshore escrow account provides the security, and exchange regulations or an instruction to pay locally hollow it out without touching the contract.
- 04An export ban produces a loss with no non-payment at all: there is no refusal to pay, there are goods immobilized.
- 05Cover must match the structure and not the family of risk: a policy bought on the wrong mechanism leaves one bare on the one that operates.