Connectedness Cycles in Equity Markets: A Wavelet Approach

The connectedness of international equity markets can be measured building on the well-established forecast error variance decomposition framework. This approach permits the assessment of the propagation of shocks (spillovers) across equity markets on a day-to-day basis. The focus of our contribution is on detecting cycles in the intensity of spillovers. As it is vital for enterprise managers to track business cycles, it is vital for investors seeking to diversify their portfolios to track the cyclicality of spillovers. Our approach provides cycle information almost in real time, while business cycles are identified not before a cycle has been completed. We apply forecast error variance decomposition in a vector autoregression (VAR) model to rolling time windows to derive indices of spillover of shocks between markets. The time series of spillover indices is then analyzed by means of continuous Morlet wavelet transforms in order to obtain dynamic insight into the spillover's composition of cycles. International connectedness has significantly increased during the last two decades, while its variability has diminished. We find patterns consisting of superpositions of cycles in the spillover series, where the prevailing cycles have recently become longer. The composition of different cycles in the spillover series has changed as well, and is now confined to a narrower band of frequencies than in the late 1980s.