An earlier module presented national content as one of the four bounds that close the public window. Here we take it for itself, because it is the only one of the four that depends entirely on the exporter and the only one it can lose without noticing. A public export credit agency supports its country's exports, not exports in general: its guarantee is conditioned on a sufficient share of the contract's value coming from the economy it serves. That condition looks administrative and is in reality a continuing obligation, verified at the least favourable moment there is, that of the loss.
The first difficulty is knowing what is counted. National share is not measured in tonnes or in tariff lines but in value, and the value of an industrial contract divides among equipment, engineering, erection, services and sometimes financing. Depending on whether one includes engineering done at head office, training, the first years' maintenance or the margin, one obtains percentages differing by ten points on the same contract. Each agency's rules say which of those elements count, and the first error is to present a figure computed on in-house intuition rather than on those rules.
The second difficulty lies in rules of origin, which do not match intuition either. A good's origin is not its maker's country nor the country it ships from: it is the country where it underwent its last substantial transformation, a notion with its own criteria. An imported component, significantly assembled in a national plant, can therefore become national; the same component merely repackaged does not. Many contracts that look insufficient on a first calculation become eligible once that question is properly asked, and the reverse is true too, which is more dangerous.
There are moreover cumulation mechanisms that widen the calculation and that exporters often ignore. Some agencies accept content from neighbouring or partner states as assimilated, under conditions and sometimes within a stated limit. Others tolerate a foreign share above the ordinary threshold where the components are unobtainable domestically, which is asked for and documented. A contractor with forty percent of value from three third countries is therefore not necessarily out of scope: it is out of scope if nobody asked the question, which is not the same thing.
Then comes the mechanism that produces the real losses, and it is organizational rather than legal. National share is measured on the contract as presented at underwriting, but an industrial contract lives for months or years: a subassembly is bought elsewhere to hold a price, a subcontractor is replaced after a failure, a production line is moved to another group plant. Each of those decisions is taken by a department with no reason to know a guarantee depends on it, and none of them reaches the finance department. The share falls silently, and nobody recalculates it until the day the agency checks.
The consequence is that a monitoring clause is worth more than a warranty clause. What protects is not a carefully drafted initial declaration but a control point attached to purchasing decisions: a threshold beyond which a change of source triggers a recalculation, a box in the subcontractor approval flow, a table kept up to date by the person following the contract rather than the one who assembled the file. As with forfeiture, what holds is not vigilance but mechanism, and the cost of that mechanism is trifling against what it protects.
That leaves what to do when the share really is insufficient, and there are three routes rather than a surrender. The first is to recut the contract, taking foreign supplies out of the guaranteed perimeter so they are contracted separately, which mechanically raises the share of the rest. The second is to approach an agency in one of the third countries concerned, whose criteria the same transaction may satisfy seen from there, or a joint guarantee between two agencies. The third is the private market, which has no such bound and charges for it. None is free, and all are decided before the review, never during it.
A French contractor signs a 52 million euro contract for a water treatment unit in Central Asia, and applies for its national agency's guarantee, whose national content threshold is 60 percent. The file presented states 68 percent. That calculation includes margin and basic engineering carried out at head office, and counts as national two pump sets manufactured outside Europe then assembled and tested at the contractor's French plant. Fourteen months after signature, the purchasing department moved manufacture of a control station to an Asian subsidiary of the group, for 3.1 million. What must be checked, and in what order?
Three questions arise and they do not carry the same weight, which sets the order. The first is the scope of the calculation: including margin and basic engineering at head office is not self-evident and is read in the agency's rules, not in in-house practice. Depending on the answer, the 68 percent stated may be worth ten points less, and one would then have been below the threshold from the outset, the worst case since the file was reviewed on that figure. The second is the origin of the pumps, and it cuts the other way: assembly and testing carried out in France may constitute substantial transformation and make those sets national, in which case the safety margin is real. Those two points must therefore be settled first, because they decide whether the contract was ever eligible. The third is the transfer of the control station, 3.1 million out of 52, that is six points of value moving from national to foreign in a purchasing decision taken fourteen months after underwriting, by a department unaware a guarantee depended on it. Taken alone it is not fatal; added to a contested calculation scope, it is. Three steps follow: have the share recalculated on the agency's written rules rather than on the initial file, document the transformation the pumps underwent while the assembly records still exist, and put a control point on purchasing decisions for the rest of performance, because verification will happen at the loss, when nothing can be corrected.
- 01National content is the only bound of the public window that depends on the exporter, and the only one it can lose without noticing.
- 02It is measured in value on the agency's rules: including or excluding head office engineering and margin moves the result by ten points.
- 03A good's origin is the country of its last substantial transformation, not its maker's: significant assembly can nationalize an imported component.
- 04Cumulation mechanisms and derogations for unobtainable components exist, and serve only those who ask.
- 05The share falls in purchasing decisions taken by departments unaware of the guarantee: a control point beats a carefully drafted initial declaration.