Jiawei Hu

About Me

I am a finance PhD candidate at the University of Texas at Dallas. I expect to graduate in 2026. My research interests lie in behavioral finance, household finance, machine learning in finance, and asset pricing. My current agenda examines the impact of behavioral biases on household financial decision-making and their subsequent consequences using machine learning methods.

Working Papaers

• Local Gambling Preference and Excessive Mortgage Leverage
  Job market paper
  Conference presentation: FMARC 2025 (doctoral), Boulder Summer Conference 2025 (poster), FMA European 2025 (doctoral; regular), IAFDS 2025 (doctoral), EFMA 2025, FMA Annual 2025 (doctoral consortium), SFA 2025
  I study the role of borrower behavioral heterogeneity in excessive mortgage leverage. Using misreported simultaneous second liens as a revealed measure of excessive mortgage leverage, I show that such borrowing was more prevalent in areas with stronger local gambling preference. The effect is stronger where borrower incentives to obtain leverage are greater, including owner-occupied and purchase loans, and is not primarily driven by lender facilitation around the FICO 620 securitization cutoff. The findings suggest that borrower behavioral traits, distinct from broad aggregate borrowing conditions, contributed to excessive mortgage leverage and financial fragility.
• Cumulative Prospect Theory and Stock Returns
  with Jun Li and Feng Zhao
  Conference presentation: FMA European 2023, EFMA 2023 (by coauthor), Eastern FA 2024 (by coauthor)
  Recent studies find that cumulative prospect theory can explain stock returns in the cross-section. We find that the explanatory power varies with the estimated probability weights from the empirical pricing kernel. Allowing time-varying probability weights strengthens the support for the cumulative prospect theory in explaining stock returns beyond existing return predictors and time-varying stock characteristics. A conditional strategy based on time-varying probability weights significantly improves the performance of the unconditional strategy.
• When Gender Disparities Meet Financial Technology in Financial Advisory
  with Xiaolin Wang, Feng Zhao, and Zhiqiang (Eric) Zheng
  Can financial technology level the playing field for female financial advisors? In the setting of copy trading on a social trading platform, we find that female lead traders attract more investors than male lead traders, controlling for performance, risk, trader characteristics, and country and time fixed effects. This finding is more pronounced among traders engaging in high-risk trading, or in countries with more women participating in the labor force. In the dynamic relation of investor flow responding to traders’ performance, we find a more convex relation between performance and flow for female lead traders, in that female traders attract more investors with good performance and lose fewer investors with poor performance than male traders. Using Baidu AI face detection, we find that older and less attractive female lead traders attract more investors. Our findings indicate that investors on the social trading platform trust female lead traders more.
• Expected Utility Theory and Prospect Theory in Skewness Preference
  Based on Barberis et al. (2021) prospect theory asset pricing model, we build a model that incorporates both expected utility theory term and prospect theory term to study skewness preference. The model is used to explain coskewness premium and idiosyn cratic skewness. As expected, prospect theory term mainly responsible for idiosyncratic skewness premium. However, we find that although both terms play a role in explain ing coskewness premium, the two terms have competing effects.

Work In Progress

• Heterogeneity in Mortgage Default: Insights from Second-Lien Misreporting
  This project studies the default behavior of primary-residence mortgage borrowers with simultaneous second liens during the housing boom, comparing those who misreported the lien with those who correctly reported it. Results show that borrowers who misreported are more liquidity-driven in default, while those who reported correctly are more equity-driven. This pattern implies differences in how borrowers value homeownership, highlighting the importance of recognizing such heterogeneity when assessing default risk. To operationalize this insight, I am developing a machine learning-based Liquidity Shock Score that predicts borrower distress out-of-sample, providing lenders with a tool that complements CLTV to better forecast defaults across heterogeneous borrower types.
• Extracting Borrower Behavioral Traits from Credit Data: A Machine Learning Approach
  Using machine learning methods, this project aims to develop borrower-level behavioral trait measures from monthly Equifax credit histories. Drawing on behavioral and household finance literature, I focus on traits such as gambling preference, overconfidence, and present bias. These traits are inferred from patterns in credit use, delinquency dynamics, and responses to local shocks. The traits are then linked to mortgage outcomes to study their role in shaping household financial performance. This work bridges behavioral finance and household finance, and highlights the potential of machine learning to extract behavioral insights from large-scale financial data.

Honors & Awards

• GARP Risk Management Award Nomination by EFMA 2025
• The Best Research Proposal at the Responsible Use of Generative AI in Research Ideation Workshop by IAFDS 2025
• The Maria Strydom Best Presenter by IAFDS 2025
• Betty and Gifford Johnson Travel Awards by UTD 2025
• UT Dallas Graduate Studies Scholarship by UTD 2020-2025

Teaching Experience

Instructor

• Business Finance (FIN 3320): Summer 2025
• Personal Finance (FIN 3300): Fall 2023, Fall 2024

Teaching Assistant

• Financial Modeling for Investment Analysis (FIN 6353): Spring 2025
• Options and Futures Markets (FIN 4340): Spring 2025
• Personal Finance (FIN 3300): Summer 2024
• Intermediate Financial Management (FIN 4310): Spring 2024
• Business in a Global World (BA1320): Spring 2023, Summer 2023
• Derivatives Markets (FIN 6360): Spring 2023, Spring 2024, Spring 2025
• Applied Econometrics and Time Series Analysis (MECO 6312): Fall 2022

Fun

On my free time, I enjoy swimming and fitness.