2025 ABFER Annual Conference (3rd place, Best Poster Award)
2025 ABFER Annual Conference (3rd place, Best Poster Award)
Firms' threat to file for bankruptcy becomes more credible with more efficient courts, so the appeal of out-of-court restructuring increases with bankruptcy efficiency. I use a large exogenous increase in bankruptcy efficiency in a triple-differences framework to verify this bankruptcy-restructuring complementarity, and show that the welfare effect is not unambiguously positive.
Using Kiva crowdfunding data, we show that prosocial cues such as smiles promote altruistic behaviors on average, and are most effective when sender-recipient cultural distances are moderately high. There is a “cultural valley”: The smile effects are the strongest when cultural distances are intermediate; they weaken when the cultures of the recipients and the senders become too similar or too distinct.
with Tse-Chun Lin
2024 Knut Wicksell Conference on Crypto and Fintech, 2024 Research in Behaviroal Finance Conference, 2024 HKU Enterprise Sustainability Conference (Best Long Paper Award), 2024 New Zealand Finance Meeting
We study the effect of crowdfunding on microfinance. Before crowdfunding, households value inclusive projects but cannot commit to pay extra for them. Crowdfunding enables households to fund these projects and eliminates the commitment problem. Using Kiva crowdfunding data, we show that the female share of microfinance borrowers increases by ten percentage points in historically less gender-inclusive regions, as result of crowdfunding partnership.
with Tse-Chun Lin
2024 World Finance & Banking Symposium
In the cryptocurrency market, liquidity is top-heavy, and short positions are constrained. After removing illiquid coins from the cross-section and making test assets long-only, the existing cryptocurrency pricing models no longer describe the cryptocurrency risk-return trade-off. We construct a mean-variance efficient (MVE) portfolio that uses network momentum to infer expected returns. The network MVE portfolio is implementable in real time, spans the existing cryptocurrency pricing models, and implies a pricing kernel that positively predicts three network growth measures over horizons up to 52 weeks. These results suggest that network theories can describe the achievable efficient frontier in the cryptocurrency market.
We conduct a difference-in-differences analysis to study the effect of marijuana use on domestic violence by exploiting variations in the enactment of retail marijuana law (RML) in Denver-Aurora-Lakewood MSA from 2011 to 2016. The RML enactment in 2014 leads to a substantial increase in domestic violence rate. We also find that RML reduces crimes that are more likely to be gang-related, such as aggravated assault and robbery.