Purpose – Based on weekly data from 2012 to 2024, this paper aims to evaluate empirically the integration and contagion properties of some emerging stock markets from North Africa including Morocco, Tunisia and Egypt, and to deepen the understanding of the linkage between them during stable and turmoil periods (Covid 19, Ukrainian war and Gazza war). Design/methodology/approach – Besides traditional Granger causality (GC) test (Granger, 1969), the (Shi, Hurn, & Phillips, 2020)’ time-varying (TV) GC test, the (Song & Taamouti, 2020)’ quantiles GC test, and the (Breitung-Candelon, 2006)’ frequency domain (FD) GC tests are used for the contagion (diversification) check between market volatility (returns). Then, the returns DCC- GARCH specifications are used for the integration investigations. Then, based on the returns DCC dynamic regressions, the contagion analysis between considered markets that are related to the unexpected events is done. Findings – As the results from the standard GC, all considered tests reveal that in mean, Tunisian returns R_T and Egyptian R_E are predictable by Moroccan R_M. Only Tunisian and Egyptian return can play then the role of diversifier. Results from these causality tests detect some contagion in variance between markets, which was denied from dynamic DDC regression regressions in returns. From dynamic DCC-GARCH model, our empirical results show a weak integration between returns. Originality/value – Via the dynamic DCC ARCH and the DCC quantile regression, the time varying GC, the quantile GC, and the spectral GC tests, this paper provides a deeper understanding of North African marginal stock market behavior and linkage.
