Dynamic beta estimation by Fama-French CAPM under Markov Regime Switches

The capital assets pricing model (CAPM) takes important place in financial theory, because it provides opportunity to describe the relationship between expected return and risk of a financial assets. There are two key restrictions in the model: first of all, there is an assumption that market risk premium is sole measure of risk, secondly that relation between risk and return is linear. One solution is using of Fama-French model under Markov Regime Switches. Currently there are no research of testing Fama-French model under Markov Regime Switches for Russian stock market. The main aim of Master paper is researching of existing of nonlinear relationship between multifactorial risk measure and expected excess stock return using Fama-French model under Markov Regime Switches on Russian stock market and determination of optimal modification free factor capital assets pricing model for Russian stock market. As for results we can conclude that using of Fama-French model under Markov Regime Switches is more effective for analyzing relationship between expected excess return and multifactorial risk measure better in compare with classic CAPM and unconditional Fama-French model. Historical regime probabilities pointed in that there are relationship between periods of state related to high factors volatility and periods than important macroeconomics events appeared. Analysis of Fama-French model parameters under Markov Regime Switches shown that module of parameters higher for state related to period of high volatility of assets return. Testing of models with different specifications resulted in conclusion that Fama-French model under Markov Regime Switches with switching of all parameters better in description of relationship between excess expected return and multifactorial risk measure.