Research and technical projects at the intersection of business questions, causal inference, machine learning, and AI.
Using data on nearly 14 million U.S. businesses and Facebook-derived social capital measures, this study applies Double Machine Learning and quantile regression to reveal that social cohesion acts as a safety net for firm survival, while economic connectedness primarily benefits top-performing firms.
The purpose of this essay was to study and understand the secondary effects of social pressure mailers on voter turnout through information diffusion.
A three-stage deep hedging architecture combining autoencoder compression, LSTM forecasting, and reinforcement learning achieves up to 46% improvement over Black-Scholes delta hedging for put options using real market data from 2011-2023.
The purpose of this essay was to study and understand the incentives of the Big Three firms to underinvest in corporate stewardship.