- Capital Regulations and the Rise of Shadow Banking, joint with Hyunju Lee and Sunyoung Lee
The Basel III regulations mandated a significant increase in the required level of banks' own capital holding. Using a new micro-level dataset of corporate credit for the largest firms in South Korea in years 2013-2019, we document a 25% decline in credit from regulated banks, and an increase of similar magnitude from non-bank lenders. We use our data to estimate a strongly negative relationship between corporate lending and minimum capital requirements. For identification, we exploit a gradual implementation of the reform in Korea and control for various confounding factors. To understand this finding, we build a quantitative model with heterogeneous banks who accumulate equity and invest in risky loans. In addition, heterogeneous firms may endogenously choose to become a non-bank lender who does not face regulations. We find that an increase in the capital requirement similar to that of Basel III can justify the decrease in regulated bank lending, as well as the rise in shadow lending of the magnitude documented in our data.
- The Ultralong Sovereign Default Risk
Between 2010 and 2015, Mexican government issued external debt worth 0.5% of its GDP in the form of century bonds. Using a sovereign default model with endogenous maturity and variable risk-free rate, I propose a theory of the ultralong debt issuances and investigate the resulting bond spreads. Government issues such bonds in order to insure against low-frequency movements in the risk-free rate, and the benefit from such hedging is largest when interest rates are low. The model calibrated to Mexico’s default history predicts lower spreads on ultralong bonds than in the data. This suggests that Mexico is expected to remain a frequent defaulter in the next 100 years.
- Optimal Taxation with Risky Human Capital and Retirement Savings, joint with Pei Cheng Yu
We study optimal tax policies with human capital investment and retirement savings for present-biased agents. Agents are heterogeneous in their innate ability and make risky education investments which determines their labor productivity and affects their consumption path. We characterize the optimal wedges and show that they are different across education groups. More specifically, we demonstrate how the optimal policy encourages human capital investment with savings incentives. We show that the optimum can be implemented with income-contingent student loans, and existing retirement policies augmented by a new tax instrument that subsidizes retirement savings for college graduates. The proposed instrument takes the form of employer’s 401(k) matching contribution proportional to the repayment of student loans, and mimics the latest policy proposals that aim to incentivize college education. We show that the optimal tax system yields significant welfare gains relative to the optimal policies designed for time-consistent agents.
- Learning about Debt Crises, R&R at AEJ: Macro
The European debt crisis presents a challenge to our understanding of the relationship between government bond yields and economic fundamentals. I argue that information frictions are an important missing element, and support that claim with evidence on the evolution of GDP forecast errors in 2008-2014. I build a quantitative model of sovereign default where output features rare disasters and agents learn about their realizations. Debt crises coincide with economic depressions and develop gradually while markets update their expectations about future income. Calibrated to Portuguese economy, the model replicates the comovement of bond spreads and output before and after 2008.
- Commitment versus Flexibility and Sticky Prices: Evidence from Life Insurance, joint with Pei Cheng Yu
Life insurance premiums display significant rigidity in the data, on average adjusting once every 3 years by more than 10%. This contrasts with the underlying marginal cost which exhibits considerable volatility due to the movements in interest and mortality rates. We build a model where policyholders are held-up by long-term insurance contracts which presents a time inconsistency problem for the firms. The optimal contract takes the form of a simple cutoff rule: premiums are rigid for cost realizations smaller than the threshold, while adjustments must be large and are only possible when cost realizations exceed it. We use a calibrated version of the model to show that it matches the data and captures several aspects of premium rigidity in the cross-section and over time.