AGUS SACCHAL sells sheets and blankets from a warehouse in Buenos Aires, for which he is paid in Argentine pesos. While the pesos go into his wallet, two other banknotes are stuck to his office window. One is a ten-yuan note from a visit to China, where he went in search of cheap textiles. The other is a $5 bill, pinned next to an invoice, also in dollars. Though he does not trade with America directly, when importing he uses the greenback.

Argentina’s rocky financial history makes the dollar’s dominance there unsurprising. Still, it is an extreme case of a wider phenomenon. After gathering data on 91% of the world’s imports, by value, Gita Gopinath of Harvard University found that America accounts for nearly 10%. But its currency is used in over 40% of invoicing.

Recent research suggests that this creates a link between a weak dollar and buoyant trade flows—and vice versa. Trends since 1999 are suggestive (see chart). During 2017 the dollar depreciated by 7% against a basket of other currencies, as global trade flows surged by 4.5%. Some other factors could be driving both. But a recent paper by Ms Gopinath, Emine Boz of the IMF and Mikkel Plagborg-Møller of Princeton University found that, even after adjusting for countries’ business cycles, a 1% dollar strengthening predicted a fall in trade volumes outside America of 0.6%.

They explain the connection by upending the standard way of thinking about the impact of exchange rates on trade. Textbook models tend to assume that importers face prices in the exporting country’s currency, which are hard to renegotiate. An importer whose currency falls against the exporter’s is squeezed. But his countrymen who export in the opposite direction get a fillip, as their wares become more competitive. In this neat and symmetric world, as a country’s imports fall because of a weaker currency, its exports rise.

But what of importers like Mr Sacchal, who buy in dollars? The researchers argue that here, the symmetry breaks down. A stronger dollar squashes his demand for Chinese products, without Argentine exporters to China gaining a countervailing bump. A strong dollar would then mean that trade volumes outside America fall.

Supporting their theory, they find that dollar exchange rates seem to be more useful than those of other currencies when predicting changes in trade flows and prices. This is particularly so in places that invoice a higher share of imports in dollars.

Alternatively, as suggested in a recent working paper published by the Bank for International Settlements, a strong dollar could tighten global credit conditions, making it harder to finance long supply chains and so crimping trade flows. The authors find that a strong dollar is associated with slower-growing company inventories (shorter supply chains require less stock to be held along the way).

Given the dollar’s recent weakness, what does all this suggest about future trade flows? The recent trade surge might be only temporary, if traders renegotiate dollar prices. The results of Ms Gopinath and her coauthors suggest otherwise. They find that, since 2002, the effects of dollar movements on trade have persisted. Gabriel Sterne of Oxford Economics, a consultancy, reckons that about half of the increase in trade flows due to the weak dollar since 2017 is yet to come.