Business Cycle and Stock Market Volatility: A Particle Filter Approach

The recent observed decline of business cycle variability suggests that broad macroeconomic risk may have fallen as well. This may in turn have some impact on equity risk premia. We investigate the latent structures in the volatilities of the business cycle and stock market valuations by estimating a Markov switching stochastic volatility model. We propose a sequential Monte Carlo technique for the Bayesian inference on both the unknown parameters and the latent variables of the hidden Markov model. Sequential importance sampling is used for filtering the latent variables and kernel estimator with a multiple-bandwidth is employed to reconstruct the parameter posterior distribution. We find that the switch to lower variability has occurred in both business cycle and stock market variables along similar patterns.