A Study of Stock Returns based on Multi-Factor Models using Panel Regression Data

Abstract: Factor models have evolved over time to study the impact of various factors on excess stock returns over time. Most factors are based on fundamentals, technical, specific strategies, or screeners. This study aims to capture a combination of these by using factors that are based on market driven factors like market capitalisation, momentum, among others, and fundamental factors like F-score, P/B ratio and others. The excess returns have been regressed against these factors to identify which of them are significant and correlated to the returns that have been achieved over a ten-year period. The findings indicate that based on the random effect model over time momentum and market cap indicators make difference in returns for the bunch of stocks chosen for the analysis. The impact of the study is to gain understanding using panel data regression about the impact on excess returns of selected factors namely; Market premium, Size, and Momentum, Book to Price ratio and F-score, and the ability to predict the returns. Keywords: BSE Sensex, Multi Factor Model, Piotroski’s F-score, Stock Returns JEL Classification Number: C2, G11, G1