Researcher ORCID Identifier


Graduation Year


Date of Submission


Document Type

Open Access Senior Thesis

Degree Name

Bachelor of Arts



Reader 1

Eric Hughson

Rights Information

2022 Jessica Cuna Zamora


With its roots dating back to 1887, Groundhog Day has cemented itself as a beloved American holiday where people gather to see if Punxsutawney Phil will predict 6 more weeks of winter, or an early arrival of spring. Utilizing the data and methodology framework from Shanaev, Savva, and Fedorova (2021) to test if Groundhog Day predictions have any effect on S&P 500 returns, this paper revisits and revises the analysis in attempt to replicate and improve the original findings with a dummy-variable regression while controlling for other calendar anomalies. Additionally, this study expands the original analysis by including two new tests: a weather effect analysis and a steel industry specific analysis. It is found that Groundhog Day predictions do not create statistically significant returns under either prediction, despite the findings of the base paper that find significant returns for early spring predictions.