Testing for financial spillovers in calm and turbulent periods

It has been established in the literature that volatility of stock returns exhibits complex properties of not only volatility clustering, but also long memory, regime change, and substantial outliers during turbulent and calm periods. Hence, this paper seeks to analyze volatility spillover, co-movements, independence and contagion in the Chinese, Japanese, Nigerian, South African, UK and US stock markets. Using a sample period spanning 2010M1 to 2018M12, the paper employs a state-space parameterization of Markov Regime Switching Model with skew-normal distribution and the Markov Chain Monte Carlo (MCMC) estimation procedure in the empirical analyses. The AMGARCH model with BEKK and DCC specifications was applied to detect presence of spillover and contagion effects from and across the financial markets as well. Posterior estimates show that the behavior of the series differ across two regimes; tranquil and crises. Specifically, market co-movements with the US Dow Jones rises during turbulence than during periods of tranquility particularly in the case of Nigeria NSE and the Johannesburg JSE. Furthermore, while the Nigeria's NSE and the JSE are entangled, the Japanese NIKKIE and the Chinese SHANGHAI show strong independence. We report evidence of bi-directional spillover transmission between the developed market (US market) and emerging markets (Nigeria and South African markets) to be asymmetrical and the likelihood for the markets to exhibit higher spillover from calm to turbulent regimes. Invariably, this validates the expectation that contagion is transmitted from the stronger markets to the weaker markets.