Clustering cryptocurrencies market through the innovative DM-MSTP method

Cryptocurrencies illustrate rapid technological transformation, market diversification, and growing adoption by investors. Clustering cryptocurrencies into homogeneous groups enables investors and portfolio managers to better understand and control risk transmission mechanisms and market co-movements, ultimately optimizing portfolio construction and enhancing risk-return management. This paper introduces a new Artificial Intelligence method, Dhouib-Matrix-MSTP (DM-MSTP), to cluster the cryptocurrencies market. At first, the correlation matrix between the whole thirty-five cryptocurrencies is converted as a distance matrix. At second, the DM-MSTP method is developed to present the minimum spanning tree joining the all thirty-five cryptocurrencies (as a topological representation). Finally and to help the decision-maker, the minimum spanning tree represented by DM-MSTP can be used to cluster the cryptocurrencies by groups.