Search Options
Home Media Explainers Research & Publications Statistics Monetary Policy The €uro Payments & Markets Careers
Suggestions
Sort by
Anja Brüggen
Georgios Georgiadis
Principal Economist · International & European Relations, International Policy Analysis
Arnaud Mehl
Adviser · International & European Relations, International Policy Analysis

Global trade invoicing patterns: new insights and the influence of geopolitics

By Anja Brüggen, Georgios Georgiadis and Arnaud Mehl

Published as part of the The international role of the euro, June 2025.

This special feature uses new data collected by ECB and IMF staff from over 120 countries to examine trends in global trade invoicing currency patterns up to 2023. The analysis reveals several key insights: the US dollar and the euro remain the most prominent primary invoicing currencies, together accounting for over 80% of global trade invoicing. While the US dollar serves as a global vehicle currency, the euro's role as a vehicle currency is particularly significant in Europe and parts of Africa. Despite some growth, the renminbi's share in global trade invoicing remains very low, at less than 2%, although it is increasing in the Asia-Pacific region and in some parts of Europe. Finally, this special feature presents some evidence of a relationship between shifts in invoicing currency patterns and geopolitical alignment, especially since Russia's full-scale invasion of Ukraine. This evidence is most marked for certain countries which have distanced themselves geopolitically from the West, such as Russia, Belarus, Kyrgyzstan and Uzbekistan, where the share of exports invoiced in the US dollars and euro was 10-50 percentage points lower in 2023 than in 2015-19.

1 Motivation and context

Trade invoicing currency choices have important implications for the transmission of shocks and policy effectiveness. The standard assumption in traditional open-economy macroeconomic models, such as the Mundell-Flemming model, is that firms set and invoice prices in the currency of the producer.[1] Over short to medium-term horizons, when prices are sticky, this pricing assumption suggests that when a country’s exchange rate depreciates, domestic spending will switch from imports to domestically produced goods and that exports will rise. Recent models depart from this assumption and postulate instead that export prices are set in a vehicle currency – i.e. the currency of neither the exporter nor the importer, but of a third country.[2] The key observation underlying this assumption – also known as the “dominant currency” pricing paradigm – is that most global trade is invoiced in just a few currencies, most often the US dollar, and sometimes the euro, regardless of the countries involved.[3] The implications of this assumption differ from those of traditional models. In particular, when export prices are set in a vehicle currency, a depreciation of the exchange rate does not boost a country’s exports. For instance, if China’s exports to Germany are priced in US dollars, a depreciation of the Chinese renminbi against the euro does not make China’s exports cheaper for German buyers, so it does not boost China’s export sales. Also, an appreciation of the vehicle currency raises the prices of all imports, not only of those sourced from the vehicle currency’s issuer. As the choice of the export pricing currency has implications for the impact of exchange rate movements on prices and quantities, it also affects the effectiveness of domestic monetary policy through the exchange rate channel and monetary policy spillovers, especially from the United States.

An open question is whether the expansion of China and other emerging market economies in global manufacturing and trade could reshape global invoicing trade patterns, which have historically centred on the US dollar and the euro. China’s spectacular rise has changed patterns in global final goods trade, input-output linkages and the structure of competition in export markets. Standard invoicing currency choice models suggest that these secular changes may make it optimal for some exporters to switch from established vehicle currencies to the renminbi. Such a shift would align their invoicing currency choice with that of foreign suppliers and competitors in destination markets increasingly using the renminbi. However, since larger switches would require coalescing effects in addition to changes in patterns in global trade in final goods, input-output linkages and the structure of competition in export markets, standard models predict that changes in invoicing currency patterns are likely to remain limited.

Additionally, policy initiatives and rising geopolitical tensions could reshape global invoicing currency patterns. Since the global financial crisis of 2009, the Chinese authorities have launched various initiatives to promote the internationalisation of the renminbi. These initiatives include allowing Chinese firms to settle trade transactions in renminbi, establishing renminbi swap lines with foreign central banks and partially opening up China’s financial account to non-residents. Providing the option and reducing the costs for settling external transactions in renminbi might influence exporters’ decisions to use it as an invoicing currency. Moreover, following the financial sanctions imposed on Russia following its full-scale invasion of Ukraine, alternative cross-border payment systems and corridors have emerged, challenging established correspondent banking networks and messaging systems. Rising and potentially prohibitive costs for settling transactions in major currencies might have also influenced exporters’ decisions regarding their choice of invoicing currency (see Section B.3 below for a more detailed discussion on the role of geopolitics in invoicing currency patterns).

Assessing whether invoicing currency patterns in global trade have shifted owing to these developments is challenging, as the relevant information is not readily available in existing cross-country datasets. Standard cross-country datasets like Comtrade, compiled by the United Nations, and the Direction of Trade Statistics, compiled by the IMF include detailed information on bilateral goods trade, but do not provide information on invoicing currency patterns.[4] So far, these patterns have not been included in the IMF’s Balance of Payments Manual – the main set of guidelines coordinating global accounting standards for external sector statistics. However, it will become a supplementary item that countries are encouraged to report from the Manual’s seventh edition onwards.

Invoicing currency information is collected from customs declarations which differ across country and time depending on national legislation. Information on invoicing currency patterns is not always a mandatory – or even a voluntary – item that importers and exporters are required to disclose. Moreover, even if the information is provided, it might not be stored or processed by customs authorities, or transferred to other authorities, such as national statistics offices or central banks.[5]

A new initiative implemented by ECB and IMF staff sheds light on developments invoicing currency patterns in global trade. Various studies have collected data on invoicing currency patterns in global trade.[6] Earlier efforts were generally confined to advanced economies, restricted to collecting information on invoicing patterns in only the US dollar and the euro. Recent work by ECB and IMF staff has significantly expanded the cross-sectional coverage of earlier studies, adding observations on a larger array of emerging and developing countries for the period from 1990 to 2019.[7] ECB and IMF staff have now updated this dataset in follow-up work, with a view to including the period following Russia's full-scale invasion of Ukraine. The updated data include additional countries and, importantly, information on invoicing currency patterns in the renminbi. This updated and expanded database provides information on the shares of exports and imports of more than 120 countries from 1990 to 2023 invoiced in the US dollar, euro, renminbi and other currencies.

2 Stylised facts and trends in global trade invoicing currency patterns

The updated and expanded data suggest that the US dollar and the euro continued to account for the lion’s share of global trade invoicing up to 2023.[8] The new data confirm key facts observed in earlier studies. Chart B.1, panel a) shows that the share of global exports invoiced in US dollars, at about 40%, remains much larger than the share of exports to the United States. As noted earlier, this difference testifies to the US dollar’s dominant role in the invoicing of global exports. Patterns for imports are similar. Chart B.1, panel b) confirms that the US dollar's leading role reflects more than its use for invoicing commodity exports: even if exports of commodities are excluded, the dollar share of invoicing still exceeds the share of exports to the United States by a sizeable margin. Chart B.1, panel a) also reveals that the euro's share of global export invoicing, at more than 40%, remains as large as the share of the US dollar, in line with the euro area’s larger trade openness relative to that of the United States.[9] The chart also shows that the euro’s share of global export invoicing is not much larger than the share of exports to euro area countries – an observation also made in earlier studies.[10]

Chart B.1

Global export shares by invoicing currency and destination

a) By invoicing currency and destination

b) By invoicing currency and destination, singling out US dollar invoicing of commodity exports

Sources: ECB staff calculations based on the analysis in Boz et al. (2025) and updated and expanded data from Boz et al. (2022), Taiwan Ministry of Finance, IMF Direction of Trade Statistics and World Development Indicators.
Notes: Following Boz et al. (2025), missing country invoicing data are interpolated and extrapolated to obtain a balanced panel. For extrapolation, the earliest (latest) available data point is used to extend backwards (forwards); in each case, the value of the first (last) available data point is held constant. Invoicing shares are rescaled proportionately if the sum across currencies exceeds 100%. Interpolated and extrapolated raw data are averaged over time from 1999 to 2023. As in Gopinath (2015), the United States is excluded from the sample. The difference between panel a) and panel b) is that panel b) singles out the invoicing of commodity exports from total invoicing in US dollars. Since data on the invoicing currency for commodities exports are not available, commodity exports are assumed to be invoiced in US dollars (even if this is not necessarily always the case).

Another salient fact confirmed in the new data is that the US dollar continues to be used as a global vehicle currency for trade invoicing, while the use of the euro is more restricted geographically. Chart B.2, panel a) shows that, for most countries, the share of exports invoiced in US dollars is a multiple of the share of exports to the United States, which reflects the dollar’s use as a global vehicle currency. By contrast, Chart B.2, panel b) shows that only countries in or close to Europe, as well as some parts of Africa, invoice a larger share of their exports in euro than their share of exports to the euro area. This points to the relevance of the euro as a regional rather than a global vehicle currency.[11]

Chart B.2

Average share of exports by destination and invoicing currency at the country level between 1990 and 2023

a) Share of exports to the United States and US dollar export invoicing share

b) Share of exports to the euro area and euro export invoicing share

(x-axis: share of exports to United States, percentages; y-axis: US dollar export invoicing share, percentages)

(x-axis: share of exports to euro area, percentages; y-axis: euro export invoicing share, percentages)

Sources: ECB staff calculations based on analysis in Boz et al. (2025) and expanded and updated data from Boz et al. (2022), Taiwan Ministry of Finance, IMF Direction of Trade Statistics and World Development Indicators.
Notes: Data are averaged over the period 1999-2023. Country names on the chart are displayed as three-letter ISO codes.

The updated and expanded data suggest that patterns in global US dollar and euro invoicing shares continue to remain remarkably stable over time, even after Russia’s full-scale invasion of Ukraine. Chart B.3 expands Chart B.1 over time. Chart B.3, panel a) shows the evolution of export shares by destination, while Chart B.3, panel b) shows the evolution of export shares by invoicing currency. Chart B.3, panel a) shows that the share of global exports to the United States and the euro area, respectively, has declined since 2000, in line with the decline in their share of global output. At the same time, Chart B.3, panel b) shows that the share of global exports invoiced in US dollars and euros has remained broadly stable. Importantly, the share of the US dollar and the euro in global trade invoicing remained stable in both 2022 and 2023 – in the wake of Russia’s invasion.

Chart B.3

Evolution of the share of global exports by invoicing currency and destination

a) By destination

b) By invoicing currency

Sources: ECB staff calculations based on analysis in Boz et al. (2025) and expanded and updated data from Boz et al. (2022), IMF Direction of Trade Statistics, Taiwan Ministry of Finance and World Development Indicators.
Notes: In Chart B.3, panel b), following Boz et al. (2025), missing country invoicing data are interpolated and extrapolated to obtain a balanced panel. For extrapolation, the earliest (latest) available data point is used to extend backwards (forwards); in each case, the value of the first (last) available data point is held constant. Invoicing shares are rescaled proportionately if the sum across currencies exceeds 100%. Estimates exclude the US but include the euro area countries, in line with Gopinath (2015). If euro area countries are also excluded, the shares of the US dollar and the euro were about 60% and 25% in 2023, respectively. The latest observation is for 2023.

Another important insight from the new data is that, while increasing, the renminbi’s share of global trade invoicing remains very low. Chart B.4, panel a) adds information on the share of global exports to China and Chart B.4, panel b) on the share of global exports invoiced in renminbi, both of which are available for a smaller sample of countries compared with Chart B.3. Although the share of global exports to China has increased significantly since it joined the World Trade Organization in 2001, the share of global exports invoiced in renminbi remains very low. There was a small uptick after 2022, but this is barely noticeable from a global perspective.[12]

Chart B.4

Evolution of the share of global exports by invoicing currency and destination (smaller sample with renminbi invoicing currency information)

a) Global exports by destination

b) Global exports by invoicing currency

Sources: ECB staff calculations based on analysis in Boz et al. (2025) and updated expanded and data from Boz et al. (2022), IMF Direction of Trade Statistics, Taiwan Ministry of Finance and World Development Indicators.
Notes: In Chart B.3, panel b), following Boz et al. (2025), missing country invoicing data are interpolated and extrapolated to obtain a balanced panel. For extrapolation, the earliest (latest) available data point is used to extend backwards (forwards); in each case, the value of the first (last) available data point is held constant. Invoicing shares are rescaled proportionately if the sum across currencies exceeds 100%. The export and invoicing currency shares are slightly different from those in Chart B.2 because of differences in the country sample, as a country is only included if data are available for all export shares and all invoicing currency shares. As in Gopinath (2015), the United States is excluded from the sample. The latest observation is for 2023.

Invoicing in renminbi has increased more rapidly in certain regions, such as Asia-Pacific and parts of Europe, but still accounts for less than 2% of global exports. Chart B.5 presents patterns in invoicing currency shares for exports in different regions. In the Asia-Pacific region, invoicing of exports in renminbi started to rise from 2010. In the rest of the world, the use of the renminbi was negligible until 2018, but has subsequently risen rapidly in parts of Europe and the Western Hemisphere. However, the share of renminbi invoicing remains below 1-2% of exports in each region.

Chart B.5

Evolution of the share of exports invoiced in renminbi by region

a) Unweighted

b) Export weighted

Sources: ECB staff calculations based on analysis in Boz et al. (2025) and updated and expanded data from Boz et al. (2022), IMF Direction of Trade Statistics, Taiwan Ministry of Finance and World Development Indicators.
Note: The latest observation is for 2023.

3 Changes in invoicing currency patterns and geopolitical alignment

Geopolitics could induce shifts in invoicing currency patterns by changing the structure of global trade networks. Standard models emphasise the role of predetermined trade patterns, input-output linkages and the structure of competition in export markets as determinants of an exporter’s choice of invoicing currency.[13] The key insight from these models is that exporters choose the currency in which their optimal price is expected to be most stable – or deviations of optimal reset from optimal preset prices are least volatile – considering future shocks to demand for their products and to the cost of their inputs. Geopolitical tensions could lead to a redirection of trade flows for final goods and intermediate inputs in cross-border value chains away from a globally integrated economy dominated by an established vehicle currency towards fragmented blocs. Fragmentation of global trade could result from changes in relative trading costs – including owing to tariffs – or even trade restrictions motivated by geopolitical considerations.[14] In blocs that are geopolitically distant and less integrated with the issuer of the established vehicle currency, exporters could be incentivised to abandon the latter for another unit of invoicing.

Other factors could also influence trade invoicing currency patterns within the existing structure of the global trade network. Standard invoicing currency choice models abstract from the role of settlement of trade transactions. In fact, in the empirical literature, the conventional assumption is that the settlement currency coincides with the invoicing currency, meaning that settlement of payments is inconsequential compared with the pricing of exports.[15] However, this might not be a valid assumption in trade between geopolitically divergent economies. For example, if its use is prohibited, or financial sanctions are (or are expected to be) imposed that reduce the availability of a certain currency to settle transactions, it may no longer be optimal to use it to invoice exports. The reason is that for such a currency it becomes increasingly difficult to minimise deviations of the optimal preset sticky price from the optimal price the exporters would choose if they could reset their price when shocks occur.[16] Interactions in international currency choice for different uses have been studied for trade pricing, financing and saving decisions, but less so for pricing and settlement decisions.[17]

Alternatives to traditional payment infrastructures centred around the US dollar and the euro have gained momentum since Russia’s full-scale invasion of Ukraine, underscoring the importance of geopolitical considerations for trade invoicing. Alternative infrastructures started to emerge in 2013 when Iran, in response to its exclusion from Swift, developed its own messaging system. Russia followed suit in 2014 with the creation of the System for Transfer of Financial Messages (SPFS) after the annexation of Crimea. China's Cross-Border Interbank Payment System (CIPS) was launched in 2015. The pace of these initiatives has accelerated significantly since Russia's invasion of Ukraine. In the past two years alone, more than 20 new initiatives have been launched by emerging countries as alternatives to traditional payment infrastructures. At the BRICS Summit in October 2024, leaders from Brazil, Russia, India, China, South Africa and other nations supported the increased use of local currencies in global financial transactions. They also discussed establishing a new cross-border settlement and depository infrastructure, BRICS Clear, to facilitate this transition.

The updated data compiled by the ECB and the IMF suggest there is weak evidence of a relationship between changes in invoicing currency patterns and geopolitical alignment after the rise in geopolitical tensions. Chart B.6, panel a) shows that across all countries in the sample, Russia, Belarus, Kyrgyzstan and Uzbekistan experienced the largest declines in the share of exports invoiced in US dollars and euro in 2023 relative to the average between 2016 and 2021. In the case of Russia, the combined invoicing share of the US dollar and euro have roughly halved.[18] Chart B.6, panel b) indicates that the declines in US dollar and euro invoicing shares for these countries coincided with an increase in geopolitical divergence vis-à-vis the United States and the imposing of sanctions on Russia and Belarus, for both trade and financial transactions.[19]

Chart B.6

Changes in the invoicing share of the US dollar and euro and geopolitical alignment between the average for 2016-21 and 2023

a) Largest declines in the combined invoicing share of the US dollar and euro

b) Correlation between changes in the invoicing share of US dollar and euro and geopolitical alignment

(percentages)

(x-axis: changes in the share of US dollar/euro invoicing in percentages; y-axis: change in ideal point distance vis-à-vis the United States, percentages)

Sources: ECB staff calculations based on analysis in Boz et al. (2025) and updated and expanded data from Boz et al. (2022), IMF Direction of Trade Statistics, Taiwan Ministry of Finance and Bailey et al. (2009).
Notes: Geopolitical alignment vis-à-vis the United States is measured as the ideal point distance derived from Voeten, E., Strezhnev, A. and Bailey, M., “United Nations general assembly voting data”, Harvard Dataverse, 32, 2009. The measure is computed using data on voting in meetings of the General Assembly of the United Nations. In panel b), euro and US dollar invoicing shares represent the average export and import invoicing shares for each country. Country names on the chart are displayed as three-letter ISO codes. The latest observation is for 2023.

The relationship between changes in invoicing currency patterns and geopolitical alignment is robust to controlling for various confounding factors. Country-level invoicing currency patterns may also change due to composition effects unrelated to changes in producer decisions about pricing currency. For instance, for given US dollar and euro pricing decisions taken by US and euro area firms, when a country increases its imports from the euro area at the expense of imports from the United States, the share of its imports invoiced in euro increases, while that of its imports in US dollars declines. Similar effects may materialise owing to exchange rate valuation effects and commodity price changes. Together, these effects can blur the relationship between changes in a firm’s invoicing currency decisions and geopolitical alignment in aggregate data and may even induce spurious correlations. However, Table B.1 shows that declines in US dollar and euro invoicing shares remain correlated with declines in geopolitical alignment even after controlling for these confounding factors. The evidence of the role of geopolitical alignment is relatively weak. Excluding Russia, Belarus, Kyrgyzstan and Uzbekistan – the four countries for which the association is the clearest (Chart B.6, panel b) – from the estimation sample, the coefficient estimate of the change in geopolitical alignment variable is no longer statistically significant.

Table B.1

Declines in US dollar and euro invoicing shares correlate to some extent with increases in geopolitical distance

Regression estimates for changes in invoicing currency shares

Imports

Exports

(1)
USD

(2)
EUR

(3)
EUR/USD

(4)
CNY

(5)
USD

(6)
EUR

(7)
EUR/USD

(8)
CNY

Change in geopolitical distance

-1.25
(0.22)

-0.84**
(0.01)

-2.26**
(0.05)

-0.24*
(0.09)

-2.48*
(0.08)

-0.67
(0.28)

-2.48
(0.12)

-0.09
(0.46)

Change in bilateral trade share

-0.36
(0.32)

0.24
(0.13)

0.34
(0.12)

0.10*
(0.09)

-0.27*
(0.10)

0.29**
(0.03)

0.35*
(0.08)

0.07
(0.33)

Change in trade block share

0.20
(0.26)

0.23**
(0.04)

0.32**
(0.05)

-0.05
(0.23)

-0.01
(0.96)

0.02
(0.89)

0.13
(0.48)

0.02
(0.49)

Change in bilateral exchange rate

0.11*
(0.08)

0.00
(0.89)

0.08
(0.28)

-0.01
(0.34)

0.14
(0.11)

0.00
(0.96)

0.20*
(0.06)

-0.03
(0.21)

R-squared

0.09

0.13

0.25

0.19

0.14

0.14

0.24

0.09

Countries

97

95

94

81

97

96

95

79

Sources: ECB staff calculations based on analysis in Boz et al. (2025) and updated and expanded data from Boz et al. (2022), IMF Direction of Trade Statistics, Taiwan Ministry of Finance and Bailey et al. (2009).
Notes: P-values are in brackets. ** denotes significance at 5% level and * significance at 10% level.

Overall, the findings suggest that global invoicing currency patterns have remained broadly stable against a backdrop of rising geopolitical tensions. This stability is consistent with models that highlight lock-in effects and network externalities as forces shaping invoicing currency patterns. These externalities are rooted in input-output linkages in cross-border value chains, strategic complementarities in price setting among exporters and domestic producers, and global exchange rate arrangements that reflect asymmetries in volatilities due to policy frameworks and other factors.

In the future, it is crucial to monitor global trade invoicing currency patterns to identify potential tipping points that could trigger more noticeable changes. State-of-the-art models suggest that large shocks and structural changes could coordinate the behaviour of producers into a different invoicing currency choice equilibrium.[20] Model simulations suggest that when the structural determinants of invoicing currency choice undergo large changes, reshuffling of global trade invoicing patterns can unfold rapidly. Moreover, rising geopolitical tensions might give fresh impetus to state policies that deliberately encourage or enforce changes in trade invoicing currency choices. Therefore, it remains important to monitor patterns in invoicing currency in global trade transactions in the period ahead.

  1. See Obstfeld, M. and Rogoff, K.S., “Foundations of International Macroeconomics”, MIT Press, 1996. The currency in which firms set optimal prices does not necessarily coincide with the invoicing currency specified on customs declarations. Evidence finding that prices are sticky in the currency of invoicing in the short term suggests that they correlate, however; see Amiti, M., Itskhki, O. and Konings, J., “Dominant Currencies: How Firms Choose Currency Invoicing and WhyIit Matters”, Quarterly Journal of Economics, Vol. 137, Issue 3, 2022, pp. 1435-1493.

  2. See Gopinath, G., Boz, E., Casas, C., Diez, F., Gourinchas, P.-O. and Plagborg-Møller, M., “Dominant Currency Paradigm”, American Economic Review, Vol. 110, No 3, 2020, pp. 677-719.

  3. See Goldberg, L. and Tille, C., “Vehicle Currency Use in International Trade”, Journal of International Economics, Vol. 76, No 2, 2008, pp. 177-192 and Gopinath, G., “The International Price System”, Working Paper, No 21646, National Bureau of Economic Research, 2015.

  4. The United Nations’ Comtrade database is maintained by the United Nations Statistics Division, which works directly with national customs authorities, statistics offices, relevant ministries and central banks to collect data on exports and imports. The IMF’s Direction of Trade Statistics relies on data reported by IMF member countries, augmented by trade statistics from other international organisations, such as Eurostat COMEXT and United Nations Comtrade for countries which do not report to the IMF.

  5. Overall, cross-country reporting and processing of trade invoicing currency information is largely uneven. Only a few countries routinely provide official data on their trade invoicing currency.

  6. See, for instance, Kamps, A., “The euro as invoicing currency in international trade”, Working Paper Series, No 665, ECB, 2006; Goldberg, L. and Tille, C., “Vehicle Currency Use in International Trade”, Journal of International Economics, Vol. 76, No 2, 2008, pp. 177-192; Ito, H. and Chinn, M., “The Rise of the ‘Redback’ and the People’s Republic of China’s Capital Account Liberalization: an Empirical Analysis of the Determinants of Invoicing Currencies”, Working Paper, No 473, Asian Development Bank Institute, 2014; and Gopinath, G., “The International Price System”, Working Paper, No 21646, National Bureau of Economic Research, 2015.

  7. Boz, E., Casas, C., Georgiadis, G., Gopinath, G., Le Mezo, H., Mehl, A. and Nguyen T., “Patterns of invoicing currency in global trade: New evidence”, Journal of International Economics, Vol. 136, Issue C, 2022.

  8. The analysis of trends in the updated and expanded data from Boz et al. (2022) is carried out in Boz, E., Brüggen, A., Casas, C., Georgiadis, G., Gopinath, G. and Mehl, A., “Patterns of invoicing currency in global trade in a fragmented world economy”, Working Paper Series, ECB, 2025, forthcoming.

  9. These estimates exclude the United States and include the euro area countries, in line with Gopinath (2015). If euro area countries are excluded, the share of the US dollar and the euro was about 60% and 25% respectively in 2023.

  10. Considering intra-euro area trade does not artifactually boost the share of global exports invoiced in euro, as intra-euro area trade is included when calculating the shares of both invoicing and trade.

  11. See also Mehl, A., Mlikota, M. and Van Robays, I., “How is a leading international currency replaced by another? Old versus new evidence”, published as a special feature in The international role of the euro, ECB, 2023.

  12. Chart B.4 does not use renminbi, US dollar or euro data on cross-border settlements by currency available for China from 2010 onwards. According to data for 2024, more than one-third of China’s exports of goods were settled in renminbi. Further analysis is needed to test whether the assumption that the currency of settlement coincides with the currency of invoicing – which is the conventional assumption in the literature (see e.g. Friberg, R. and Wilander, F., “The currency denomination of exports – a questionnaire study”, Journal of International Economics, Vol. 75, Issue 1, 2008, pp. 54-69 – is valid for China’s trade. If data on China’s settlements were added in the chart, the share of the renminbi in global trade would be substantially larger.

  13. See, for instance, Engel, C., “Equivalence Results for Optimal Pass-Through, Optimal Indexing to Exchange Rates, and Optimal Choice of Currency for Export Pricing”, Journal of European Economic Association, Vol. 4, Issue 6, 2006, pp. 1249-60; Gopinath, G., Itskhoki, O. and Rigobon, R., “Currency Choice and Exchange Rate Pass-Through”, American Economic Review, Vol. 100, No 1, 2010, pp. 304-336; and Mukhin, D., “An Equilibrium Model of the International Price System”, American Economic Review, Vol. 112, Issue 2, 2022, pp. 650-688.

  14. See Clayton, C., Maggiori, M. and. Schreger, J., “A Theory of Economic Coercion and Fragmentation”, Working Paper, No 33309, National Bureau of Economic Research, 2024; Trebesch, C., Meyer, J., Zhou Wu, J., Martin, A. and Broner, F., “Hegemony and International Alignment”, Working Paper, No 1483, Barcelona School of Economics, 2025; and Martin, P., Mayer, T. and Thoenig, M., “Make Trade Not War?”, Review of Economic Studies, Vol. 75, Issue 3, 2008, pp. 865-900.

  15. See Friberg and Wilander (2008), op. cit.

  16. Specifically, when settlement in the invoicing currency, e.g. the US dollar, is prohibited, an exporting firm may simply quote its price set in US dollars in its home currency. In this case, the exporter would be exposed to exchange rate risk from the moment it sets and quotes the price of its output to the moment it sells and receives payment for its output. If the home (i.e. the quoting and settlement) currency happens to depreciate against the US dollar (i.e. the price setting currency) between pricing/quoting and selling, the preset invoicing currency price in US dollars and the home currency will turn out to have been too low. If the movement in the exchange rate had been predictable, the exporter could have factored in the depreciation and set a higher price in US dollars and hence quoted a higher price in its home currency. But since exchange rates are hard to forecast and may also appreciate, quoting and settling in the home currency implies more volatile deviations of optimal reset from optimal preset US dollar prices, which undermines the attractiveness of the US dollar as an invoicing currency.

  17. See Chahrour, R. and Valchev, R., “Trade Finance and the Durability of the Dollar”, Review of Economic Studies, Vol. 89, Issue 4, 2021, pp. 1873-1910; and Gopinath, G. and Stein, J., “Banking, Trade, and the Making of a Dominant Currency”, Quarterly Journal of Economics, Vol. 136, Issue 2, 2020, pp. 783-830.

  18. For reasons of availability, in Russia, Belarus, Kazakhstan and Kyrgyzstan, the data refer to trade settlement rather than trade invoicing as specified on customs declarations. This is noteworthy because evidence suggests that the Bank of Russia has drawn from its swap line with the People’s Bank of China to provide renminbi liquidity to Russian commercial banks to enable them to pay for imports from China and save US dollar reserves; see Horn, S., Parks, B., Reinhart, C. and Trebesch, C., “China as an International Lender of Last Resort”, NBER Working Paper, No 31105, 2023.

  19. Geopolitical alignment vis-à-vis the United States is measured as the ideal point distance measure put forward by Voeten, E., Strezhnev, A. and Bailey, M., “United Nations general assembly voting data”, Harvard Dataverse, 32, 2009. This measure is computed using data on voting in meetings of the General Assembly of the United Nations.

  20. Calibrated simulations of Mukhin (2022)’s rich state-of-the-art general equilibrium models suggest that further increases in China and other emerging market economies’ share of global trade and output, combined with the liberalisation of the renminbi exchange rate and economies choosing the renminbi as an anchor currency, would not challenge the dominant role of the US dollar for invoicing global trade. However, the simulations also suggest that a structural increase in US inflation to 10% due to fiscal policy slippages could induce economies to abandon the US dollar as an anchor currency and look for alternatives, thereby significantly reducing its role in the invoicing of global trade.