Siwon Lee

Evaluates Earnings-at-Risk models under trade policy uncertainty, separating tail misspecification from firm heterogeneity using quantile and conformal methods, with calibration by firm size.

Extends a Merton portfolio-choice model to include trade-policy uncertainty, jump risk, and stochastic equity volatility, and approximates its HJB solution operator with a multiple-input DeepONet.

Estimates household stock, bond, and cash portfolio responses to trade-policy uncertainty shocks, and models transmission through firm inventories and imperfect inventory information. The paper develops a Merton-style framework for robust portfolio choice under Knightian uncertainty.

Studies firm financial vulnerability after exchange-rate depreciation, including heterogeneity by trade openness and financial hedging. It uses Deep Equilibrium Networks to solve nonconvex equilibrium conditions and develops a structural model with endogenous exchange rates and heterogeneous firms.

Models correlated automated trading, price feedback, and leveraged liquidation in continuous time, and analyzes position limits, margin requirements, and circuit breakers with reopening auctions across India, the US, and the EU.

Compares multilingual risk disclosures of US-listed Korean and Japanese firms using language models, studies cross-language differences in disclosure, and develops tests to detect and mitigate measurement bias in multilingual text analysis.

Uses agent-based simulations of limit-order-book trading to study liquidity and volatility responses to large shareholder sales, including how information asymmetry, risk preferences, and forecast feedback affect price impact.

Builds a state-year panel of labor-market indicators and mortgage delinquency to study links between labor-market vulnerability and household financial distress, including lagged specifications and robustness checks.