Unlocking Market Secrets: Statistical Factor Model in Stock Return Prediction

This paper delves into the domain of statistical factor models as a unique approach for constructing factor-based investment strategies with the goal of outperforming the stock market. Unlike conventional factor models that explicitly identify and quantify the factors influencing stock returns, the statistical factor model operates with latent and concealed factors, rendering it akin to a 'black box' strategy.The primary allure of the statistical factor model lies in its potential to unearth and utilize hidden factors that may remain unnoticed through traditional analysis. This approach allows investors to potentially tap into novel sources of return, enhancing their portfolio performance. However, this unorthodox characteristic also presents a challenge: the lack of transparency regarding the precise factors driving returns. The study tests the statistical factor approach on the constituents of NSE 500 index from 2013 to 2023 and found that the model comprehensively beats NSE 500 buy and hold strategy.