Research
(Please note: I don't disclose early-stage projects)
(Please note: I don't disclose early-stage projects)
Abstract: This paper documents partisan sorting between issuers and lead underwriters in U.S. initial public offerings. Using a deal-level dataset spanning 1990--2024 that links LSEG records to BoardEx executive and Federal Election Commission donations, I show that politically divergent issuer--underwriter pairs are systematically less likely to match. A conditional logit selection model implies that a one-unit increase in political divergence reduces selection probability by approximately 8 percentage points---about 35% of the unconditional match rate. The result is causally identified using the staggered geographic expansion of the Sinclair Broadcast Group as a plausibly exogenous shock to underwriter executive ideology, and operates through trust and information-asymmetry channels: the effect intensifies during polarized periods and in informationally opaque deals, and attenuates when external certification substitutes for partisan-mediated trust. Partisan divergence also carries pricing implications: a one-standard-deviation increase in deal-level political divergence is associated with 6.4 percentage points higher first-day underpricing.
Presentations: 2026 Southern Finance Association (Scheduled); 2026 FMA (Scheduled); 2026 European Financial Management Association (Scheduled); Iowa Finance Brown Bag Seminar; 2026 SouthWestern Finance Association
with Amrita Nain and Yi Hao
Abstract: We show that ideological disagreement among sell-side analysts generates persistent earnings forecast dispersion. This effect is stronger in politically contentious industries and for firms with high ESG scores. An immigration-based shift-share instrument and a novel firm-level measure of analyst ideological disagreement confirm causality. Unlike information-based dispersion, ideology-induced dispersion does not decline over time. Consequently, ideological disagreement has significant implications for asset pricing and corporate finance. Stocks covered by more ideologically polarized analysts earn lower future returns, and acquirers with polarized analyst coverage use more equity and, in the subset of politically contentious industries, earn lower announcement returns.
Presentations: 2025 SFS Cavalcade North America 2025*; 2025 Eastern Finance Association Annual Meeting; 2025 FMA Annual Meeting; 2025 Silicon Prairie Finance Conference; 2024 Spring & Fall Iowa Finance Brown Bag Seminar*; University of Texas at San Antonio*;
Awards: 2025 FMA Best Paper Semi-finalist in Asset Pricing & Investment
Under Review
with Jiajie Xu
Abstract: We study how political distance between places affects the mobility of skilled labor, the diffusion of ideas, and corporate innovation. In a county-pair panel of U.S. inventor mobility from 2001 to 2023, a one-standard-deviation increase in political distance between origin and destination counties reduces bilateral inventor flows by 17%, within as well as across party lines. At the firm level, a one-standard-deviation increase in PAC-donation-based ideological distance implies a 62.5% decline in inter-firm inventor moves. A regression discontinuity design around close presidential elections identifies a pull mechanism: destinations that narrowly elect Republicans attract fewer inventors, with no comparable origin-side effect. Political distance also dampens knowledge diffusion, reducing bilateral patent citations by 9.6%, slows entry into new technology classes, and lowers patenting output, both across counties and within firms' own footprints. Partisan polarization thus imposes a real cost on the allocation of inventive talent.
Presentations: 2026 FMA (Scheduled); 2026 Silicon Prairie Finance Conference
with Yunxin Yi
Abstract: We examine whether the partisan composition of an analyst's workplace affects her earnings forecast accuracy. Matching sell-side analysts in IBES to voter registration records over 2000--2024, we find that analysts at more politically diverse brokerages produce significantly more accurate forecasts, even after absorbing firm-by-year, analyst, and brokerage fixed effects. Comparing theoretical extremes, the most diverse workplaces are associated with approximately 3.9% lower absolute forecast error than the most homogeneous. Restricting to analysts who switch brokerages yields larger estimates, consistent with a causal workplace effect. Event-study analyses around five U.S. presidential elections show that the accuracy advantage of diverse workplaces attenuates sharply in post-election months while homogeneous workplaces are unaffected, suggesting that diversity benefits forecast quality through informal information exchange that is disrupted when partisan identities become salient.
with Amrita Nain, Tong Yao
with Amrita Nain, Yasmine Nosair, and Jiajie Xu
Abstract: We study whether funding goes up for firms developing labor-saving automation patents facing COVID-induced labor supply shock. We look into different types of funding, including venture capital funding, corporate acquisitions, and government funding such as SBIR and SBA. Additionally, we examine whether VC experience matters in identifying the firms conducting labor-saving innovation as well as the value of such innovation. Employing a work-from-home suitability measure, we find that innovative firms that are shielded from the labor shock themselves and develop more labor-saving innovations attract more VC funding, particularly experienced VCs. We find no similar effects for corporate acquisitions and government funding.
Presentations: 2024 Spring Finance Brown Bag Seminar