Author
In times of economic stress, households, businesses, and governments around the world seek stable places to invest. Since at least the close of World War II, this global “flight to safety” has led mostly to the United States. While recent trade and fiscal policy developments may have taken some shine off the dollar, U.S. assets—especially the deep, liquid market for U.S. Treasury debt—still hold a distinctive position as a safe haven.
It might seem surprising, then, that this well-known global flight-to-safety (GFS) phenomenon has not played an explicit role in workhorse economic models. In recently updated research, Minneapolis Fed Research Director Andrea Raffo and former colleagues from the Federal Reserve Board of Governors find that integrating GFS shocks into a standard macroeconomic model can not only explain patterns of global economic and financial variables but also reconcile macroeconomic data with the observed fluctuations in currency exchange rates (Minneapolis Fed Working Paper 799, “Global Flight to Safety, Business Cycles, and the Dollar” with Bodenstein et al.).1
This second finding addresses a long-running concern that macroeconomic models are poor at explaining and forecasting the relative values of currencies based on other fundamental economic data—an “exchange rate puzzle.” Richard Meese and Kenneth Rogoff, for example, first demonstrated in 1983 that a “random walk” model provided a better forecast of real dollar exchange rates than various structural models.
Raffo and co-authors integrate GFS shocks into a two-country (U.S. and the rest of the world) dynamic stochastic general equilibrium model estimated using Bayesian methods and 21 macrofinancial time series over the sample period from 1992 to 2019. The economists model GFS shocks as exogenous, global changes in households’ preference for safe bonds. The economists do not preordain that these bonds are denominated in U.S. dollars, although this result emerges endogenously within the model when estimated with a suite of U.S. and global macroeconomic data and standard parameters from the literature.
In one key exercise, this approach lets the economists decompose the fluctuating value of the dollar into core components (see figure).
Unsurprisingly, the dollar’s value is strongly associated with the interest rate differential between U.S. rates and a composite of the rest of the world. However, the global preference for safe assets adds substantial explanatory power to the model. GFS shocks explain about 10 percent of exchange rate fluctuations during this 30-year period, playing a prominent role around global economic slowdowns.
Adding GFS shocks greatly reduces the explanatory weight that has been typically assigned to a residual term, deviations from “uncovered interest parity” (UIP). Collectively, the result is that movements in the relative value of the dollar are largely accounted for by fundamental economic factors (resolving the “puzzle” in the literature).
Their model with GFS shocks is also much better at forecasting the dollar exchange rate over the medium run, providing superior predictions to a random walk at three- and five-year horizons. “This finding is remarkable,” the economists write, demonstrating that exchange rates can be pinned down in the model “through a useful accounting of its fundamental forces, while it also allows for policy and counterfactual analysis.”
The flight to safety is even more closely associated with other macroeconomic outcomes. During their study period of 1992 to 2019, GFS shocks explain about 22 percent of world GDP growth and 75 percent of observed borrowing spreads. GFS shocks also help explain world consumption and investment growth and (to a lesser degree) world inflation. The findings suggest changes in global risks matter greatly for world economic outcomes. This stands in contrast to theories that U.S. economic activity largely drives outcomes for foreign countries (notably, recent research by Silvia Miranda-Agrippino and Hélène Rey focused on the spillovers of U.S. monetary policy).
In both the U.S. and abroad, flights to safety contribute to a stronger dollar, wider borrowing spreads (resulting in tighter private sector and sovereign credit markets), lower consumption and investment, and weaker labor markets. In the U.S., the upward pressure on the dollar depresses exports and leads to a larger drop in inflation.
When testing alternate specifications of global shocks in their model, Raffo and co-authors confirm that a model featuring GFS shocks is a superior fit to recent global economic data. “Overall,” they write, “these statistical results confirm that using a global shock that acts like a risk premium shock biased towards dollar-denominated assets is a promising approach for explaining international comovement and relative prices over the business cycle.”
In addition to making international economic models more tractable for policy analysis, the work underscores and quantifies the global forces buffeting national economies. The global search for safe assets strongly influences economic activity and currencies in ways that can inform domestic policy decisions and might, the authors note, motivate greater policy coordination across governments.
Read the Minneapolis Fed working paper: “Global Flight to Safety, Business Cycles, and the Dollar”
Endnote
1 Board of Governors co-authors: Martin Bodenstein, Pablo Cuba Borda, Nils Gornemann, Ignacio Presno, Andrea Prestipino, and Albert Queralto.
Jeff Horwich is the senior economics writer for the Minneapolis Fed. He has been an economic journalist with public radio, commissioned examiner for the Consumer Financial Protection Bureau, and director of policy and communications for the Minneapolis Public Housing Authority. He received his master’s degree in applied economics from the University of Minnesota.
