A genetic-search model for first-day returns using IPO fundamentals

In this paper, we present a study of genetic-based stock selection models using the data of fundamentals of initial public offerings (IPOs). The stock selection model intends to derive the relative quality of the IPOs in order to obtain their relative rankings. Top-ranked IPOs can be selected to form a portfolio. In this study, we also employ Genetic Algorithms (GA) for optimization of model parameters and feature selection for input variables to the stock selection model. We will show that our proposed models deliver above-average first-day returns. Based upon the promising results obtained, we expect our GA-based methodology to advance the research in soft computing for computational finance and provide an effective solution to stock selection for IPOs in practice.