(SSRN)
Abstract: We study how data ownership affects lending markets when data production is endogenous. We distinguish portable data, which borrowers can transfer across lenders, from non-portable data, which remains proprietary to incumbent lenders. In our model, an incumbent bank chooses portable data production before a technologically advanced fintech decides whether to enter. Shifting ownership to borrowers weakens incentives to produce portable data and can lower allocative efficiency despite greater competition. Borrower ownership can jointly expand credit availability and improve allocative efficiency in some cases. Portable data production varies non-monotonically and exhibits discrete jumps as the bank's ability to deter entry changes.
Presentations:
FMA (2026); Edinburgh Financial Technology Conference (2026); 37th Stony Brook International Conference on Game Theory (2026); MRS International Risk Conference (2026); CES China Annual Conference (2026); 8th Future of Financial Information Conference (2026); Finance Theory Group Summer School (2023)
Award:
Asian FA Best Doctoral Student Paper Award (2026); Graduate Dean’s Dissertation Fellowship (2026)
Abstract: How do investor heterogeneity in size and signal quality distribution jointly amplify asset price fluctuations and market fragility? We answer that a large investor with low signal quality would be more likely to trade with the winds and would be harder to absorb the market shock, and thus create market fragilities. We develop a model where investors of varying sizes and signal qualities strategically submit asset demands in response to endogenous price impact. We investigate how this behavior absorbs or amplifies informative and uninformative shocks. Our findings suggest that while large investors typically make markets more inelastic and amplify shocks, strong information advantages can offset this effect. Using U.S. corporate bond mutual fund data (2010–2024), we simulate and identify two different policy regimes. When passive share is low, enhancing competition among informed investors could reduce liquidity-shock fragility and promote efficiency. Conversely, when passive share is high, directing high-quality information to large, active investors is more effective for market stability and price efficiency.
Presentations:
Asian Finance Association (2026); CES China Annual Conference (2026); Southwest Finance Association (2026); Eastern Finance Association (2026); AFA PhD Student Poster Session (2026); Finance Theory Group Summer School (2025); UC-Irvine PhD Research Fest (2025)
Award:
Peter H. Stevens Jr. Memorial Scholarship (2026)
Abstrct: How does generative AI affect investor behavior and asset prices? We develop a model in which AI refines public signals, while investor skill determines how strongly trades respond to those signals. Widespread AI adoption, therefore, increases commonality in trading and can alter market efficiency. We connect this mechanism to the rise of passive investing. Because passive strategies seek to replicate the market, they reinforce the price impact of public signals when the share of uninformed investors is high. Consequently, as uninformed investors increasingly rely on AI-refined signals, equilibrium prices become more sensitive to public information and less responsive to dispersed private information and other shocks. Although AI improves individual signal precision, common reliance on its output crowds out private-information aggregation, amplifies coordinated trading, and reduces price informativeness
Presentations: FMA “New Ideas” Session (2025)
Award:
Ray Watson Fellowship (2025)
Abstract: Changes in age structure across different locations coinciding with a lack of housing affordability have recently led to public discussions over the ability of young households to acquire ``starter homes'' if older households who own a substantial share of the housing stock choose to ``age in place.'' We develop a shift-share IV approach based on long-run age structures to estimate the causal effect of demand changes on house prices in different housing segment by size. We find that the housing supply elasticity is lower in cities with high senior shares, particularly for large homes. To quantify the impact of demographic structure with regard to age on housing market affordability, we develop a quantitative spatial equilibrium model that allows for changes in demand for different housing types over the lifecycle in partially segmented housing markets. We aim to show whether long-run changes in age structure can affect housing affordability.
Presentations: AREUEA National Conference (2025);
Presentations: UCIrvine, Finance Ph.D. and Ph.D. Alumni Conference (2024);
"How does the aging population affect technology innovation?", The Journal of World Economics, 2017(4), with Dongmin Yao, Jing Ning
"Upgrading? — The empirical study for the migrant worker workplace", Nankai Economics Studies, 2015(1), with Dongmin Yao