Dynamic Beta Estimation and Time-Varying Risk Premium: Evidence from NSE Companies Using CAPM Extensions

Abstract

This paper will analyze the performance of 20 actively traded NSE companies using both the static and dynamic CAPM models, and will show that time-varying beta, rolling regression, and macroeconomic sensitivities are effective in estimating risk. The computation of dynamic beta is calculated using a rolling-window regression of the daily closing prices from 2015 to 2024. Sharpe and Treynor ratios are then used to evaluate the risk-adjusted performance. The most important macroeconomic variables are included to investigate their impact on risk premiums and include interest rates, inflation, and market volatility. The findings indicate that the beta varies greatly during bull, bear, and highly volatile periods, proving the inefficiency of constant-beta models. Dynamic CAPM offers better explanatory value (Adj. R 2 = 0.72 vs. 0.41), which allows making more accurate forecasts of the expected returns. The paper also emphasizes the use of financial information systems to process high-frequency market data to help support the data-driven, adaptive portfolio strategies. In general, the results indicate the relevance of dynamic and macro-sensitive modelling to the development of enhanced financial decisions in the emerging markets.