Abstract
This paper investigates the dynamical relationship between conventional and Islamic stock markets using the wavelet-assisted cross-spectral, cross-correlation and causality analyses. Relying on bivariate time series from emerging and developed markets, the aim is to find and recognize local microscopic signs of convergence or divergence. The data set covers a period of exceptional instability in the financial system that was accompanied by a significant slump in the global economic environment. The empirical results demonstrate an obvious strong dependence between conventional and Islamic indexes at low-frequency, while the dependence becomes rather instable in the finest frequencies across different investment time horizons. The relationship also took a special different form in the crisis period compared to relatively calm periods. In developed markets, indexes were the most correlated over many periods and at many frequencies, while the relationship in emerging markets tended to be less manifest, especially for short-term horizons, offering investors different investment alternatives and portfolio diversification opportunities. The pre- and post-crisis causality investigations at the end of the study suggested a bidirectional relationship in most cases, thereby offering further perspectives on multivariate forecasting.
•The relationship between conventional and Islamic stock markets is examined using MRA.•The aim is to locate similarities and differences across different time-scales.•The analysis revealed unusual patterns especially occurring at crises periods.•A bidirectional causality is found, thus opening new prospects for joint forecasting.