Risk Modeling on the BRVM

Abstract

Classical risk measurement models assume normality of returns and linear dependence, which systematically underestimates extreme losses in emerging markets. This paper evaluates two complementary approaches to model the risk of a portfolio of BRVM assets (SONATURG and BOA BF) over 2020–2023. The first uses copulas to capture tail dependence between assets. The second employs Markov chains to model market regime persistence. Results show that the Student's t-copula captures non-zero tail dependence (0.28) and the Markov chain reveals strong regime inertia (probability of staying in same regime above 0.85). Combining both approaches provides a more comprehensive view of risk for BRVM investors.