How Do Regulations and Technology Affect Service Allocation and Market Structure?

Under review

Selected for 13th World Congress of the Econometric Society (ESWC 2025), Seoul.
Presented at Asia Meeting of the Econometric Society (AMES) 2024 and Eastern Economic Association (CSWEP session) 2024.

Where Has the Foreign Signal Gone? (with Ulrich Hounyo)

Presented at North American Summer Meeting of the Econometric Society (NASM 2026), Atlanta.

Abstract: Social media sentiment predicts stock losses and gains, yet for which stocks and why remains unclear. We propose that the answer lies in the match between who discusses a stock on Twitter and who holds genuine informational advantages about it. English-language Twitter discourse is concentrated among U.S.-based investors who are well-positioned to assess domestic firms but comparatively disadvantaged in evaluating foreign stocks listed on U.S. exchanges. This mismatch predicts that U.S. Twitter sentiment should be a more reliable signal for domestic stocks than for foreign ones. Using a large panel of firm-quarter observations, we find that sentiment uncertainty predicts subsequent losses for domestic stocks but leaves foreign losses largely unpredicted. This predictive gap is present for the most typical stocks, widens for stocks of intermediate benchmark similarity, and collapses for the least typical stocks, where domestic and foreign sentiment become equally informative. Structural estimation provides evidence that the domestic advantage reflects genuine signal precision rather than mechanical differences in return volatility. These findings link the home bias literature to the informativeness of crowd-sourced signals and suggest that the value of social media data is geographically structured.

Medicaid Expansion: The Key that Opens the Job Lock?