Risk-Return Relationship: What Can a Time-Varying Approach Add? An Analysis of Alternative Modelling Techniques

The risk-return relationship has been a fundamental concept in finance, highlighting the trade-off between potential gains and the likelihood of losses. There are various model to establish the risk return relationship such as CAPM, French and Fama three factors and five factors models, etc. However, these traditional models assume a constant risk-return relationship, which may not capture the dynamic nature of financial markets. This study aims to explore the fitting and forecast gain of employing time-varying approaches in analysing the risk-return relationship. Alternative modelling techniques including Rolling Regression models, Markov-switching models, and Kalma filter-based state-space models to investigate the time-varying nature of risk and return relationship. These techniques allow for the estimation of varying risk and return parameters across different market conditions and economic regimes. The empirical investigation reveals that the risk-return relationship is not constant over time, suggesting the presence of market fluctuations. By accounting for these variations, investors and policymakers can make more informed decisions regarding asset allocation, risk management, and investment strategies, and emphasize the significance of employing time-varying approaches in analysing the risk-return relationship. The findings have implications for portfolio managers, policymakers, and investors.