• Kaaveri Ashwin

This study investigates how financial and macroeconomic indicators predict Initial Public Offer (IPO) under/overpricing (UOP) using machine learning (Random Forest, XGBoost) and traditional regression. While descriptive statistics revealed skewed, non-normal data with outliers, multicollinearity was acceptable. Traditional linear regression showed poor fit (R² = 0.25) with no significant predictors. XGBoost offered a slight improvement (R² = 0.26), identifying Interest Rate (IR) as the key predictor, followed by Return on Net Worth (RONW) and Operating Profit Ratio (OPR). Random Forest performed poorly (R² = -0.46). Overall, the study found limited linear relationships and moderate improvement with nonlinear models, stressing the need for better data quality, preprocessing, and feature selection in future research to enhance prediction accuracy.

UOP, XGBoost, financial performance, interest rate, profitability, feature importance

"MODELLING FIRM PERFORMANCE: AI POWERED ASSESSMENT OF UNDER/OVER PRICING OF INITIAL PUBLIC OFFER", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.12, Issue 7, page no.101-105, July-2025, Available :http://www.jetir.org/papers/JETIRGZ06021.pdf

"MODELLING FIRM PERFORMANCE: AI POWERED ASSESSMENT OF UNDER/OVER PRICING OF INITIAL PUBLIC OFFER", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.12, Issue 7, page no. pp101-105, July-2025, Available at : http://www.jetir.org/papers/JETIRGZ06021.pdf