Moroccan Stock Exchange market topology in crisis and non-crisis periods

Abstract

This paper seeks to investigate the dynamics within the Moroccan Stock Exchange (MSE) market topology in crisis and non-crisis periods using daily historical log returns of sectoral indices covering the period from January 4, 1993 to September 9, 2021. The study applies the Agglomerative Hierarchical Clustering (AHC) implemented on the Dynamic Time Warping (DTW) distance matrix over ten sub-periods covering numerous crises, from Subprime mortgage crisis to European debt crisis and finally COVID-19 crisis. The obtained clustering results are gathered into a network to display the cumulated interconnections between the sectoral indices. The findings showed that the Casablanca Stock Exchange (CSE) market clusters composition is dynamic during the studied period. Indeed, some sectoral indices demonstrated evidence of strong similarities by gathering in the same cluster over numerous sub-periods as the couples Electrical & Electronic Equipment and Transport or as Banks and Construction & Building Materials sectoral indices. Moreover, the interconnections of CSE sectoral indices are trend dependent. According to the obtained network, the Oil and Gas demonstrated its centrality.