Measuring Pricing Efficiency of Bank Equity by the Convergence of Fundamentals and Price-based ROE Predictions Applying Gradient Boosting

This study investigates semi-strong form market efficiency in Indian banking sector by examining the alignment between Return on Equity (ROE) forecasts based on company fundamentals and those inferred from stock price movements. Using Gradient Boosting models, two prediction frameworks are developed-one utilizing 24 fundamental financial variables and the other incorporating stock returns. The analysis spans 21 banks, measuring predictive accuracy through $\mathrm{R}^{2}$, RMSE, and efficiency ratios. Results reveal that most banks exhibit high predictive accuracy, with ROE forecasts closely aligning with stock price movements. However, deviations are observed in some banks, driven by investor sentiment and short-term fluctuations. The study highlights cognitive biases, market sentiment, and information lag as critical factors causing misalignment. The findings contribute to understanding market efficiency, providing insights for investors, policymakers, and financial analysts in enhancing equity valuation and investment decisions.