TMS package employs a Bayesian approach to test for mean reversion of the data that follows the two-regimes Markov Switching model with mixture Gaussian. The Bayesian framework is designed to draw from the posterior of the unobserved states of volatility and model parameters via Gibbs sampler. The posterior Variance Ratio(VR) test statistics of standardized observation can be generated to draw conclusion on the mean reversion
TMS package contributes to the STAT 840(Winter2020) course project, the report provides the underlying methodology of the package.