Research

I document a strong story-statistics gap in stock market misreaction to corporate news: whereas qualitative announcements exhibit mild overreaction, quantitative announcements generate pronounced underreaction. This pattern cannot be fully explained by signal-strength uncertainty, diagnostic expectations, news sentiment, or investor attention. Motivated by experimental evidence on selective recall, I hypothesize that memory evaluates similarity based on different characteristics for stories and for statistics. I propose that stories are similar within each news category due to shared narrative features, while statistics with comparable stock price impact are similar across news categories due to their numerical precision. Model-predicted misreactions under my hypothesized similarity patterns strongly align with the empirically observed story-statistics gap across news categories. I additionally find strong, measurement error-free support for my model in announcement-level data. In line with the hypothesized similarity structure, I show that news announcements that are more similar in quantitativeness generate more comparable stock price impacts.

Using TAQ data around trading-hour SEC filings from 2003 to 2024, we document a pre-announcement drift across EDGAR acceptance timestamps: when firms experience significant cash-flow news, 30-minute pre-acceptance returns positively predict 30-minute post-acceptance returns. Momentum, investor attention, recent firm-specific news, and contemporaneous market-wide movements do not explain the pattern; and this pattern would disappear if genuine filing date-times are replaced by placebo with random date-times. Extreme pre-acceptance returns coincide with elevated abnormal volume, informed-trading risk, and directional order flow, and the drift strengthens with SEC enforcement caseload. Our evidence is consistent with filing information reaching prices before public dissemination.

I provide new evidence on the value-relevance of international trade development, the heterogeneous distribution of foreign economic benefits among market participants, and the value-add of additional geographic information disclosure by designing the Trade Momentum Index with publicly available citation share, export volume, and trade barrier data. Using a sample of 13,016 firm-year combinations of goods-exporting U.S. firms between 2008 and 2020, I find that a Trade Momentum Index-based, equal-weight hedge portfolio generates a statistically significant annualized alpha of 17.42% at a Sharpe ratio of 0.8255. This result exhibits robustness as the abnormal returns persist under different factor models.

Empirical evidence on the economic consequence of information asymmetry on the performance of small and young firms is scarce. Using novel datasets on project outcomes and subsequent performance for small U.S. construction firms, I test implications of information asymmetry and identify the causal effect of past failure on future success. I find that abnormal-weather-driven failures cause small contractors to receive 9.1 less construction permits over the ensuing two years. The negative effect of past failure on average decreases with firm age. Following comparable failures, performance on average deteriorates 30% more for new firms than it does for an average-age firm.

I establish a theoretical framework of multi-stage production, vertical specialization, and contractual rigidity that quantitatively illustrates the importance of trade financing to the emergence of global value chains through the interruption of risk transmission and augmentation in supply chain networks. After calibrating model parameters with previous studies, I numerically estimate that the positive externality generated across global value chains by producers’ decision to secure insurance-like trade financing arrangements may account for 1.26% to 1.37% of the chains’ total revenue. Furthermore, I find that trade financing arrangements may be overpriced by up to 32.65% in today’s market.