We study the interaction between the currency choice of private domestic contracts and optimal monetary policy. The optimal currency choice depends on the price risk of each currency, as well as on the covariance of its price and the relative consumption needs of the agents signing the contract. When a larger share of contracts is denominated in local currency, the government can use inflation more effectively to either redistribute resources or reduce default costs, which makes local currency more attractive for private contracts. When governments lack commitment, competitive equilibria can be constrained inefficient, thus providing a reason to regulate the currency choice of private contracts. We show that both the equilibrium use of local currency and the implications for regulation depend on the level of domestic policy risk. Our model can explain the wide use of the US dollar in international trade contracts and the observed hysteresis in dollarization.

The US government is the dominant supplier of global safe assets and faces a downward sloping demand for its debt. In this paper, we ask if the US exercises its market power when issuing debt, and we study its macroeconomic consequences. We develop a model of the global economy in which US public debt generates a nonpecuniary value for its holders, analyze the equilibrium in which the US government is themonopoly provider of this safe asset, and contrast this casewith the one inwhich the US government acts as a price taker. We use variation in estimated demand elasticities for US debt during highand low-volatility regimes to empirically distinguish between these two models and find that the data reject the price-taking behavior in favor of the monopoly one. We then quantify the distortions due to market power and find that it generates a significant underprovision of safe assets, a sizable markup in the convenience yield, and large welfare benefits for the US to the detriment of the rest of the world. Finally, we study the implications of increasing competition in safe assets from other sovereigns and private institutions. ∗Aprevious draft of this paper circulated under the title “TheMacroeconomic Implications of USMarket Power in Safe Assets.” We thank Manuel Amador, Andy Atkeson, Anmol Bhandari, V. V. Chari, Chris Conlon, Marco Duarte, Simon Gilchrist, Oleg Itskhoki, Rohan Kekre, Arvind Krishnamurthy, Zhengyang Jiang, Ricardo Lagos, Hyunju Lee, Hanno Lustig, Matteo Maggiori, Lorenzo Magnolfi, Dmitry Mukhin, Chris Sullivan, and Venky Venkateswaran, as well as numerous seminar and conference participants, for helpful comments and suggestions. We also thank Duong Dang for superb research assistance.

We study the role of global financial intermediaries in international lending. We construct a model of the world economy, in which heterogeneous borrowers issue risky securities purchased by financial intermediaries. Aggregate shocks transmit internationally through financial intermediaries' net worth. The strength of this transmission is governed by the degree of frictions intermediaries face in financing their risky investments. We provide direct empirical evidence on this mechanism showing that around Lehman Brothers' bankruptcy, emerging‐market bonds held by more distressed global banks experienced larger price contractions. A quantitative analysis of the model shows that global financial intermediaries play a relevant role in driving borrowing‐cost and consumption fluctuations in emerging‐market economies, during both debt crises and regular business cycles. The portfolio of financial intermediaries and the distribution of bond holdings in the world economy are key to determine aggregate dynamics.

We study crises characterized by large adjustments of aggregate consumption through their microlevel patterns. We document the cross-sectional patterns of consumption adjustment across the income distribution and find large adjustments for top-income households, who exhibit consumption-income elasticities similar to or larger than the average. We construct a heterogeneous-agent open economy model of consumption under income fluctuations and show that the data patterns are largely consistent with theories that attribute the dynamics of aggregate consumption to changes in aggregate permanent income. We also discuss our findings’ implications for theories based on the tightening of households’ borrowing constraints and analyze policy implications.