Research

Published works

Sovereign Default and Capital Controls. 2021. |  Slides. | Appendix. | Code. (Review of International Economics)

This paper explores a role for capital control policies in enhancing a sovereign’s commitment to repay its debts. I study the equilibrium of an economy in which a sovereign must finance some expenditure, is constrained by the savings decisions of domestic households, and cannot discriminate between foreign and domestic lenders. I show that capital controls are crucial for implementing an equilibrium with lending from abroad when domestic disposable income is low. The distortion controls affect on bond prices crowds in domestic lending, enforcing repayment. In an environment with uncertainty over the cost of default, the sovereign exploits the commitment device that capital controls afford to mitigate, but not eliminate, default risk due to the distortionary cost of controls and the option value of default. The paper offers a novel rationale for countercyclical capital controls, distinct from conventional theories that stipulate controls be employed during expansions to regulate capital inflows. The optimal policy is characterized and empirical implications are discussed.

Working Papers

Consumption Behavior Across the Distribution of Liquid Assets. 2023. | Slides. | Appendix. |Extended Appendices.  (Revise and Resubmit, American Economic Journal: Macroeconomics)

I study household consumption responses to predictable income using transaction data from a U.S. financial institution. I document large consumption responses that are highly front-loaded to income receipt, decline moderately in levels of liquidity, and are significant for households with substantial liquid assets; in contrast with canonical buffer-stock theory. To interpret these facts, I develop a model of mental accounts in which households partition their consumption choice set between a current income and a current asset account. The model nests the buffer-stock and hand-to-mouth consumption models as limiting cases. I estimate the model and show that these two extremes are inconsistent with the timing and magnitude of the documented consumption responses. I show that an intermediate case, in which households are moderately averse to dissaving, predicts consumption responses across levels of liquidity that are consistent with the data. The sensitivity of households to income fluctuations has direct positive implications for the design of fiscal stimulus policies. The model predicts a redistributive stimulus to liquidity-constrained households is 53% less effective relative to a standard buffer-stock economy.