Where Will Global Financial Fragmentation Lead?

Şebnem Kalemli-Özcan, Professor of Economics at Brown University and Director of the Global Linkages Lab, is a former senior policy adviser at the International Monetary Fund and former lead economist for the Middle East and North Africa at the World Bank.

Şebnem Kalemli-Özcan 

The global monetary order is fragmenting. Each use of financial sanctions by the United States raises the option value of an alternative to the dollar, making diversification a form of strategic insurance. But while managed diversification is healthy, a disorderly scramble for the exits would not be.

Europe learned a version of this lesson in 2010, and its experience remains the best evidence we have about what currency unification can and cannot deliver. Since debates about geopolitical fragmentation always revive the late economist Robert Mundell’s dream of a world currency—or its modern variant of two leading currencies, the dollar and the euro—we should be mindful of the euro’s real-world experience over the past quarter-century.

The euro crisis was the perfect test of the optimal currency area (OCA) theory that Mundell had presented in his 1961 paper, because his criteria—factor mobility (especially that of labor), fiscal transfers, and symmetric shocks—turned out to be exactly the dimensions along which the eurozone would be challenged. A monetary union without a fiscal union and labor mobility, facing asymmetric shocks, behaved precisely as the theory predicted: it transmitted stress it could not absorb. The crisis thus seemed to vindicate Mundell’s OCA theory.

Of course, OCA criteria are not fixed in time, but rather are shaped by economic integration itself. In a 2001 paper, my colleagues and I showed that regions and countries with more specialized production structures have output fluctuations that are less correlated with everyone else’s. Combined with our earlier finding that capital-market integration causes such specialization, we concluded that financial integration pushes shocks toward greater asymmetry.

We then described the mechanism behind this pattern in a 2003 paper, showing that the more a group can share risk—across German regions, US states, or EU countries, for example—the more its members can afford to specialize and trade. Thus, insurance buys specialization, which in turn buys trade and output asymmetry. The empirical upshot of Mundell’s theory is that regions within federations share risk heavily and specialize extensively, whereas sovereign countries share almost no risk at all. The euro, on this reading, still cannot be an OCA. It had the integration that drives specialization and asymmetry, but it lacked the federal insurance that makes asymmetry survivable.

Without such insurance, a single currency misallocates capital, leading to declining productivity. As we show in a 2017 paper, the interest-rate convergence that accompanied the euro’s arrival did send a flood of cheap capital into Spain, Italy, and Portugal—exactly the “downhill” flow the textbook promised. But the textbook also predicted that this capital would find its most productive uses, and it did not—a misallocation story. In economies with size-dependent financial frictions, the falling cost of capital drew investment toward firms with high net worth rather than high productivity. As the dispersion of returns to capital across firms widened, total factor productivity fell.

This pattern appears in Spain, Italy, and Portugal, but notably not in Germany, France, or Norway, where financial markets are deeper. The euro did not merely expose its members to asymmetric shocks they could not insure against; the capital it attracted was systematically misallocated, dragging down the productivity of the periphery.

So, these mechanisms explain why monetary union has proved so much harder in practice than its architects hoped. Could the world nonetheless converge to a durable dollar-euro duopoly? Current trends suggest not. Geopolitical fragmentation is pushing the system toward many currency blocs, not toward one or two central banks. The political logic of the moment favors assertions of monetary sovereignty of every nation, not its surrender.

A generalized version of the European story is also the story of the past four decades of financial globalization. The textbook case for globalization in the 1990s promised that capital would flow downhill from rich economies to poorer ones, equalizing returns and accelerating convergence. Instead, the paradox that Robert Lucas had observed in 1990 held. Instead of capital flowing from developed to developing countries, China’s savings flowed to the US, producing the persistent imbalances and domestic grievances that now drive US trade policy. Likewise, integration was supposed to let countries insure one another against shocks. Instead, consumption remained less correlated than output across countries. It was the reverse of what efficient risk sharing predicts. The world got the contagion without the insurance.

What about the current digital-currency revolution? A naive reading casts it as a global-single-currency enabler; but, in practice, the opposite is happening. Stablecoins, the fastest-growing form of cross-border digital money, are roughly 97% dollar-denominated, and US legislation now deliberately channels digital-dollar activity into privately issued Treasury-backed tokens, with the explicit strategic aim of entrenching the dollar. As a response, China’s e-CNY—and perhaps a future digital euro—is being built as an instrument of national monetary sovereignty, for the express purpose of creating payment rails that can operate outside the dollar system. The technology that could in principle have delivered Mundell’s single world money is instead deepening dollar dominance and fortifying national monies. It is fragmenting, not unifying, the monetary order. It is only a matter of time before every country pushes for its own digital currency.

Mundell’s wished-for destination is unreachable because the world is nowhere near an OCA: labor does not move freely across borders; there is no global fiscal authority to transfer resources from booming regions to slumping ones; and shocks are profoundly asymmetric across countries. 

A single global currency would do to the world what the euro did to its periphery. It would drive further specialization and more asymmetry over time, with no fiscal union to insure against it. The man who gave us the tools to evaluate OCAs also gave us the decisive argument against the single world currency.

The international monetary system today is best understood not as Mundell’s map of regions choosing exchange-rate regimes, but as a network of trade and financial linkages through which monetary influence propagates. Resilience in such a system will come neither from a single global currency nor from a scramble into national fortresses, but from the deliberate management of the linkages that constitute the network.

Copyright: Project Syndicate, 2026.
www.project-syndicate.org