UROP Project

Machine learning and text analysis in finance: Analyse corporate disclosures of publicly traded companies

We study what information firms disclose to the public markets and how investors react towards it. We use financial and behavioural theory and empirical analysis to assess under which circumstances investors are not able to correctly interpret the disclosed information. We do so to understand which circumstances introduce mispricing and hence, an investment opportunity to the market. With our new project, we analyse how investors of special purpose acquisition companies invest based on the information disclosed around an acquisition by applying text analysis and machine learning.

Task

The main task is to develop/research a theory that can first, link financial disclosures to financial decisions of a company, and second, predict investor behaviour towards these disclosures. Subsequently, we use a large dataset to examine our hypotheses empirically and develop statistical as well as machine learning models.

Requirements

A strong background in business or economics is highly beneficial. The student should also be proficient in using Excel and doing desk research. Basic knowledge of the statistic software (e.g. Python, R, STATA, or SPSS) is desirable but not necessary.