A decade after the global financial crisis, the dollar continues to maintain its status as the chief international currency. Possible alternatives such as the euro or renminbi lack the broad financial markets that the U.S. possesses, and in the case of China the financial openness that allows foreign investors to enter and exit at will. Any change in the dollar’s predominance, therefore, will likely occur in response to geopolitical factors.

Linda S. Goldberg and Robert Lerman of the Federal Reserve Bank of New York provide an update on the dollar’s various roles. The dollar remains the dominant reserve currency, with a 63% share of global foreign exchange reserves, and serves as the anchor currency for about 65% of those countries with fixed exchange rates. The dollar is also widely utilized for private international transactions. It is used for the invoicing of 40% of the imports of countries other than the U.S., and about half of all cross-border bank claims are denominated in dollars.

This wide use of the dollar gives the U.S. government the ability to fund an increasing debt burden at relatively low interest rates. Moreover, as pointed out by the New York Times, the Trump administration can enforce its sanctions on countries such as Iran and Venezuela because global banks cannot function without access to dollars. While European leaders resent this dependence, they have yet to evolve a financial system that could serve as a viable alternative.

The dollar’s continued predominance may also reflect other factors. Barry Eichengreen of UC-Berkeley and Arnaud J. Mehl and Livia Chitu of the European Central Bank have examined the effect of geopolitical factors—the “Mars hypothesis”—versus pecuniary factors—the “Mercury hypothesis”—in determining the currency composition of the international reserves of 19 countries during the period of 1890-1913. Official reserves during this time could be held in the form of British sterling, French francs, German marks, U.S. dollars and Dutch guilders.

The authors find evidence that both sets of factors played roles. For example, a military alliance between a reserve issuing country and one that held reserves would boost the share of the currency of the reserve issuer by almost 30% if there was a military alliance between these nations. They conjecture that the reserve issuer may have used security guarantees to obtain financing from the security-dependent nation, or to serve the role of financial center when the allied country needed to borrow internationally.

Eichengreen, Mehl and Chitu then use their parameter estimates to measure by how much the dollar share of the international reserves of nations that currently have security arrangements with the U.S. would fall if such arrangements no longer existed. South Korea, for example, currently holds 84% of its foreign reserves in dollars; this share would fall to 54% in the absence of its security alliance with the U.S. Similarly, the dollar component of German foreign exchange reserves would decline from 98% to 68%.

In previous eras, such calculations might be seen as interesting only for providing counterfactuals. But the Trump administration seems intent on cutting back on America’s foreign military commitments. The U.S. and Korea, for example, have not negotiated a renewal of the Special Measures Agreement to finance the placement of U.S. troops in Korea.  German Chancellor Angela Merkel has defended her country’s role in NATO in the face of criticism from President Trump that Germany must spend more on defense expenditures. The possibility of a pan-European army to serve as an alternative security guarantee is no longer seen as totally far-fetched.

The dollar may be safe from replacement on economic grounds. But the imminent shrinkage of the British financial sector due to the United Kingdom’s withdrawal from the European Union  shows that political decisions follow their own logic, sometimes without regard for the economic consequences. If the dollar lose some of its dominance, it may be because of self-inflicted wounds.