Today we are pleased to present to you a guest post written by This is Hero (Portland State University).
There is no doubt that the US dollar is the most dominant international currency in many aspects of international finance. It has the largest market share in international trade settlement and purchase, sale and inventory, investment and financial settlement, foreign exchange market transactions, etc. In addition, more than 60% of the foreign exchange reserves of central banks around the world are in US dollars. According to Ito and Kawai (2024), 35-38% of global GDP uses the US dollar as the anchor currency. To date, no currency has emerged that can compete with the U.S. dollar.
The dominance of the dollar in the international monetary system has created a world in which shock waves are felt both for the United States, the dollar's issuer, and for other countries. Spillover effects from the United States affect global asset markets, bank lending, and economic fluctuations. In such a system, the non-U.S. world follows U.S. monetary policy decisions and other economic and financial news as much as its own policies and economic data.
While the dollar's dominance is clear, measuring the size of a currency's dominance zone is a rather tedious task. The dominant estimation methods are the subject of academic debate.
A recent paper by Ito and Kawai (2024) develops a new method for estimating the size of the major currency areas: the US dollar (USD), the euro (EUR), the Japanese yen (JPY), the British pound (GBP), and the Chinese yuan (RMB). This paper adopts a simple econometric approach to estimate the size of major currency areas, first popularized by Frankel-Wei (1994) and further developed by Kawai-Pontines (2016).
Specifically, the depreciation rate of a country's currency X against a certain digital currency (such as the New Zealand dollar) is regressed on the depreciation rates of five major currencies. This estimate yields estimated coefficients for each major currency. By performing this estimate in a scrolling window, it is possible to obtain each major currency's share over time. If a currency is pegged to the U.S. dollar (such as the Hong Kong dollar), its estimated coefficient against the U.S. dollar is 1, while other major currencies have estimated coefficients of zero. For example, if the estimated coefficient is (USD, EUR, GBP, JPY, RMB) = (0.40, 0.40, 0.05, 0.05, 0.10), then the estimated weight of the US dollar is 40% and the estimated weight of the RMB is 40%. 5% each for euro, pound, yen, 10% for renminbi, etc.
This approach has two inherent disadvantages. First, if some major currencies are correlated with each other, the estimated weights are not statistically accurate. Until recently, the yuan was explicitly or implicitly pegged to the yuan, which could lead to biased estimates. Secondly, many researchers use estimated coefficients as the weights of major currencies in currency baskets, but usually do not consider the statistical significance of the estimation results, so there is a risk of overestimating the size of the currency area. In particular, some argue that the size of the RMB zone has expanded in recent years. However, there does not appear to be any paper discussing whether the estimated coefficients required to estimate the size of this currency area are statistically significant. In other words, the size of the RMB (or other) area may be overestimated. In other words, the RMB zone may be much smaller than many estimate.
Ito and Kawai (forthcoming RIETI working paper) carefully address both of these issues. In response to the problem of potential high correlation between major currencies, they adopted the Kawai and Pontines (2016) method to reduce the bias caused by the potential high correlation between major currencies.
For the second question, this paper considers the degree of ERS, defined by the root mean square error (RMSE) of the estimated model. A higher level of RMSE indicates a lower level of fitness and therefore a lower level of ERS (more flexible exchange rate regime). Multiply the estimated currency weights by the ERS, such that if the ERS level is higher (i.e., the RMSE is lower and the exchange rate is more stable), the weight will be higher, while a higher RMSE means that the ERS is lower (i.e., the exchange rate is more flexible). system). With this modification, overestimation of the size of specific major currency areas (such as the RMB area) can be avoided. In other words, the method of Ito and Kawai (2024) determines the extent to which a country belongs to each major currency area and how statistically significant this is. Therefore, this approach avoids overestimating the size of currency areas and necessarily displays a more flexible exchange rate regime. By specifying the exchange rate regime in more detail, including not only the major currencies to which the domestic currency is stable, but also the degree of “tightness” (or “looseness”) that should be statistically considered, this approach could provide an approach that is more consistent with international currencies. The current status of the system.
Figure 1 provides a snapshot of the evolution of exchange rate regimes over the past 50 years by focusing on the anchor currency and the degree of exchange rate stability (or flexibility) of each economy. Each economy on the world map is colored based on the anchor currency with the statistically significant highest estimated weight among the major currencies. For example, in the 1975 world map, some economies (including Canada, Colombia, Indonesia, Mexico, Nigeria, and Thailand) are colored dark blue because of the highest estimated dollar weights and small RMSE levels (or large ERS indices ).
figure 1: The evolution of the major currency areas. source: Compiled by the authors based on their estimates.
In the figure, each color is colored according to the RMSE scale, which is divided into three goodness-of-fit ranges. Economies with a smaller RMSE (or a higher degree of ERS) are shown in darker colors, while economies with a larger RMSE (or a lower degree of ERS) are shown in a lighter color.
Plotting each economy with a different color density adds nuance to the analysis. Many researchers who implement the Frankel-Wei or Kawai-Pontines methods fail to incorporate fitness into the information. In other words, their approach does not clarify whether the regression results have sufficiently high explanatory power.
Figure 1 contains several interesting points. First, the U.S. dollar has been the most dominant anchor currency over the past five years. After the collapse of the Bretton Woods system in 1973, major developed economies moved to flexible exchange rate systems, but many emerging and developing economies (except for some countries that pegged their exchange rates to the currencies of the former colonial powers) decided to continue with flexible exchange rates. to stabilize its exchange rate against the US dollar. In the early 1990s, many former Soviet republics began to adopt the U.S. dollar as their anchor currency.
Second, the euro (or DEM before 1999) consolidated its position in Western Europe and spread eastwards in the 1990s and 2000s. Economies in West and Central Africa whose currencies were pegged to the franc began to choose the euro as their exchange rate anchor. However, outside the euro area and its periphery and West and Central Africa, euro dominance is not observed. Its reach is unmatched by the U.S. dollar.
Third, the number of economies that have used the pound and/or yen as anchor currencies over the past five years has been limited. By the mid-1970s, the number of economies with stable exchange rates against the pound had dwindled. As of 1975, only Guyana, India, Ireland and Sierra Leone appear to have given the pound the highest weight among major currencies. As of 2021, few economies use it in this way.
The yen also lacks a sphere of influence of its own. In 1985, when Japan's economy was at its peak, nearly 30 economies (including Iran, Myanmar, Romania, Samoa, Singapore and Sweden) at least partially stabilized their currencies against the yen. Since then, the role of the yen as an anchor currency has declined. In 2020 and 2021, about 20 economies and 7 economies respectively use the yen as a partial anchor currency.
Fourth, although in this analysis China has been considered a major currency country since 1999, the map shows only a handful of economies that are part of the RMB zone. Recently, many researchers have identified a number of economies that belong to the RMB zone. However, most of these economies are less stable towards the RMB exchange rate, as can be seen from the weaker explanatory power of the estimates (i.e. higher RMSE values). As of 2021, multiple economies (including Australia, Botswana, Brazil, Colombia, Indonesia, Russia and Uruguay) have been identified as having the highest weight of the RMB as an anchor currency among major currencies. However, the root mean square error of these economies is relatively high, so it is judged that the correlation between their currencies and the RMB is not close. If goodness of fit is not considered, economies with highly flexible exchange rates such as Brazil and Russia may be classified as RMB zone economies.
Ito and Kawai (forthcoming) also examine what factors lead or prevent countries from belonging to these five currency areas. This exercise tests two hypotheses. The first hypothesis is that the weight of a basket of currencies is affected not only by the structural characteristics of the economy, but also by trade, investment and financial links with major reserve currency countries and regions (i.e. the United States, the Euro). regions, the UK, Japan and China). The second assumption is that the weight of a major currency in a basket of currencies is determined by the major currency's share of other financial transactions. For example, the U.S. dollar's weight is not only affected by the U.S. dollar's trade, investment, and financial connections with major reserve currency countries and regions, but also by the U.S. dollar's participation in other types of financial assets.
Preliminary empirical analysis yields several interesting results.
First, the dollar weight is positively affected by the share of trade with the United States and the share of the dollar in export invoices and cross-border bank liabilities. Likewise, the euro weight is positively affected by the economy's share of trade with the euro area and the euro's share in export invoices, inward foreign direct investment stocks and cross-border bank liabilities. The yuan weight will not be significantly affected by the share of trade with China, foreign direct investment inflows or borrowings from China. All in all, there are few statistically significant determinants of RMB weighting.
These results provide evidence for the existence of network externalities. That is, for example, countries with a higher share of the U.S. dollar in trade, investment, and cross-border financial transactions tend to have higher U.S. dollar weightings in their currency baskets, even if the country in question does not share close economic and financial ties with the United States. Because of these externalities and network effects, the U.S. dollar will continue to play a leading role in the international monetary system.
refer to:
Frankl, J., and SJ Wei. 1994. Yen bloc or dollar bloc? Exchange rate policies in East Asian economies.exist Macroeconomic linkages: savings, exchange rates, and capital flows, edited by T. Ito and A. Krueger. 295–329. Chicago: University of Chicago Press.
Ito, H. and M. Kawai. coming soon. “The Size of Major Currency Areas and Their Determinants,” RIETI Working Paper Series #XXX. RIETI: Tokyo.
Kawai, Masahiro, and Victor Pontis. 2016. “Is there really a RMB bloc in Asia? An improved Frankel-Wei method.” international journal of money and finance62 (April), pp. 101-1 72–97.
