Research

Job Market Paper

Macroprudential Policy and Precautionary Deleveraging: Evidence from a Quasi-Experiment in Korea (co-authored with Hyundo Joo.) 

The latest version of the working paper can be downloaded here

What happens when homeowners unexpectedly lose access to housing collateral due to macroprudential policy tightening? This study explores this previously understudied aspect of the policy by investigating how existing homeowners in Korea respond to stricter mortgage constraints. Surprisingly, homeowners voluntarily reduce their debt as a precautionary measure, even though they can retain their original mortgage contracts. Exploiting recent regional tightening episodes of Loan-to-Value (LTV) and Debt-to-Income (DTI) limits with Korean consumer credit panel data, we find that affected homeowners reduce their debt by an additional 5-10\% of their annual income compared to unaffected groups, with the effect being particularly pronounced among highly indebted individuals. While this self-imposed deleveraging seems to enhance financial stability, the policy also increases short-term default frequency, suggesting that these protective measures may inadvertently expose homeowners to new financial vulnerabilities. Our simulations indicate that policy designs incorporating homeowner liquidity could achieve the intended debt reduction without triggering such unintended consequences.

Work in progress

Risk Accumulation in Bank Mortgage during housing boom 

Credit expansion becomes more risky during asset booms because lending standards often loosen as asset prices rise, increasing the potential for defaults if the boom reverses. Using HMDA mortgage application data and census tract-level housing price data, I demonstrate that banks more exposed to housing appreciation during the recent housing boom lowered their mortgage lending standards. This suggests that mortgage supply undergoes laxer screening during asset booms.

Demographics and Housing Cycle (coauthored with Yungu Cho, Jinho Kim and Hyunseo Park) slides

We investigate how population age structure shapes the size of housing price booms and busts. Using country-level and US district-level data, we show that areas with younger populations experience more pronounced housing price cycles. We further provide evidence that younger households exhibit more substantial delayed overreaction in their housing expectations. The Bayesian expectation model explains that the less rigid belief in the persistent shocks could drive such expectation behavior.

Macro-prudential dampening of Monetary policy 

Refinancing plays an important role in monetary policy effects on consumption. Response to the monetary policy can be dampened when households lose refinancing options because of mortgage limits. Using Dodd-Frank Act's Ability-to-pay cut-off at 43% debt-payment-to-income ratio, I study if the regions with a greater share of affected households showed weaker response to the monetary expansion.