Abstract:  The surge in foreign currency (FC) corporate debt in emerging economies has sparked concerns about macroeconomic stability, heightened by speculation about non-financial firms engaging in carry trades. Using firm-level data on the currency denomination of both assets and liabilities, we find evidence of firms’ carry trades: firms save in local currency liquid assets and earn higher interest income after issuing short-term FC debt. They also set aside FC liquid assets as FX risk buffers. A large degree of heterogeneity in incentives is observed. Notably, listed firms participate more in carry trades and allocate less FX risk buffers than non-listed firms.

Presented at:  Midwest Macro Fall 2022, Society for Economic Dynamics 2023

First version:  November 2021  

Previously circulated as "Carry Trades and Precautionary Saving: The Use of Proceeds from Foreign Currency Debt Issuance"

Abstract: Empirical work finds that flows of investments from the U.S. and other high income countries to emerging markets increase during times of quantitative easing by the U.S. Federal Reserve, and the reverse movement occurs under quantitative tightening. We offer new evidence to confirm these findings, and then propose a theory based on the liquidity of U.S. government liabilities held by the public. We hypothesize that QE, by increasing liquidity, offers greater flexibility for investors that might be concerned their funds will be tied up when shocks to income or investment opportunities arise. With the assurance that some of their portfolio can be readily sold in liquid markets, rich country investors are more willing to increase investments in illiquid loans to emerging markets. The effect of increasing the liquidity of U.S. government liabilities on investments in EMs may even be stronger during times of greater uncertainty.

Presented at: Central Bank of Brazil (scheduled)

First version: May 2024

Abstract: This paper investigates the role of dollar debt and firm heterogeneity in shaping the exchange rate pass-through to global trades. With a unique dataset that merges extensive firm-level balance sheet information with transaction-level Korean customs data, we find that firms with greater exposure to foreign currency debt tend to decrease their export quantities and increase their export prices following the devaluation at the end of 1997, particularly pronounced for smaller firms. On the other hand, larger exporters increase their export quantities and lower their export prices as they are more indebted in foreign currency. The heterogeneous price and quantity responses across firm size potentially arise from large firms, indebted in foreign currency, not facing the disruption in their production as much as smaller firms. On top of that, larger firms may increase their exports to generate more cashflows when they are more indebted in foreign currency as their production capacity is not restricted after the devaluation. We also find that exporters, indebted in foreign currency, reduce their sales to domestic markets more than domestic firms, especially when exporters are small facing tighter financial constraints. The panel data analysis from 2001-2020 confirms the relevance of the financial channel of dollar debt in the exchange rate pass-through to export prices and quantities in more recent periods.

Presented at: UNIST, North American Summer Meeting 2024, KER International Conference 2024, Asian Meeting of the Econometric Society in China 2024 (scheduled)

First version: June 2024

Abstract:  We explore the negative balance sheet effect of foreign currency borrowing on the exchange rate pass-through to domestic prices. Exploiting a large unexpected devaluation episode in Korea in 1997, we show that firms with higher foreign currency debt have indeed experienced balance sheet deterioration and faced lower growth rates of sales and net worths and reduced their price-cost markups. We then empirically document that a sector populated by firms with higher foreign currency debt exposure prior to the crisis experienced a larger price increase. Building a heterogeneous firm model with financial constraints, we quantify the role of foreign currency liabilities in explaining the exchange rate pass-through to prices and find that 20% to 80% of the sectoral price changes during the crisis can be explained by the balance sheet effect of foreign currency debt alone. We emphasize the role of strategic complementarity in amplifying the sectoral price increase.

Presented at:  Midwest Macro Fall 2019, Korea International Economic Association 2022 Winter Conference at SNU, ASSA Meeting 2023, Ohio State University, Yonsei University, University of Seoul,  NBER East Asian Seminar on Economics 2023, Washington Area International Finance Symposium, 3rd WE_ARE_IN Macroeconomics and Finance 2023,  4th Women in International Economics Conference, Johns Hopkins University - Economics Department, ASSA Meeting 2024, NBER International Finance and Macroeconomics Program Meeting Spring 2024, WE_ARE Virtual Seminar Series, Southern Economic Association 2024 (scheduled)

First version: August 2019

Abstract: Borrowing in foreign currency has historically induced a large currency mismatch on emerging economies' balance sheets, leading to financial instability and economic crises. Nonetheless, emerging market sovereigns still borrow a substantial amount in foreign currency. In fact, in this paper, we find empirically that emerging market sovereigns borrow even more in foreign currency when exchange rate volatility is higher, precisely when it is riskier for them to do so. This paper builds a quantitative sovereign default model with a risk-averse sovereign and risk-averse international investors, where the optimal currency composition of external sovereign borrowing is the outcome of a risk-sharing problem between the borrower and lenders. Emerging economies choose to bear exchange rate risk and borrow in foreign currency because international investors charge a high exchange rate risk premium on emerging market local currency debt. Moreover, the required premium on local currency debt is higher when exchange rate volatility increases, further dissuading emerging economies from borrowing in local currency. The estimated model with high risk aversion of lenders quantitatively matches well the foreign exchange risk premium, the relative borrowing cost in local currency over foreign currency, and the currency composition of external public debt. The model also performs well quantitatively in accounting for positive comovements between the foreign currency share of external public debt and exchange rate volatility, and the relative borrowing cost in local currency over foreign currency and exchange rate volatility. A counterfactual exercise shows that exchange rate stabilization results in a welfare gain to the emerging market sovereign of 0.35% measured in consumption equivalents.

Presented at:  KU Leuven Summer Event 2022,  Asian Meeting of the Econometric Society in China 2022 (AMES), Yonsei University, Society for Economic Dynamics 2022 (SED), Korea-America Economic Association, Midwest Macro Fall 2022, University of Maryland, North American Summer Meeting 2023, Sovereign Debt Workshop at the IMF 

First Version: November 2021

Abstract: This paper helps unravel the long-standing equity home bias puzzle by building a model in which an agent infrequently adjusts her portfolio holdings of home and foreign equities. As real exchange rate returns are volatile, an investor who invests in foreign equities and holds on to her portfolio holdings for a long duration is likely to drift away from an optimal allocation. The agent, taking infrequent adjustment into account ex-ante, lowers her demand for foreign equities, generating home bias in equities. The introduction of the euro into various European countries and the enlargement of euro area in subsequent years provide a natural environment in which to validate the implications of the model. We empirically document that European countries experience lower equity home bias after adopting the euro as cross-border equity investment within the euro area entails no nominal exchange rate risk. When the levels of real exchange rate volatility are calibrated to match the average levels for European countries in the euro area and outside the euro area, the model can match the difference in levels of equity home bias between European countries experienced after the introduction of the euro.

Presented at:  North American Summer Meeting  2019 (NASMES 2019), Midwest Macro Fall 2019, Economics Graduate Student Conference 2019 in St.Louis (EGSC 2019),  Western Economic Association International 2021 (Virtual, WEAI 2021)

First version:  December 2018