Inter-Market Sentiment Analysis Using Markov Switching Bayesian VAR Analysis

Abstract

This study examines the nonlinear interdependency among the volatility indexes of gold, oil, and stock markets. The volatility indexes are used as proxies for market sentiment for the period from March 2010 to March 2017. The Markov switching Bayesian vector autoregressive (MS–BVAR) method is applied to measure the interdependency of the lags of these volatility indexes. The empirical results show three unidirectional causal relationships, lag dependencies, and positive impacts of the different market sentiments. There is always a moderate volatility period between every transition from recession to expansion of volatility situations which consistently last for 41 days; high-risk and low-risk periods last for one and three day(s), respectively. The greatest impact is from the first lag of the stock market volatility on the gold market volatility. © 2022, The Author(s), under exclusive license to Springer Nature Switzerland AG.