Rethinking the CAPM beta on emerging markets: A wavelet coherence approach to the Moroccan financial sector

This study examines whether the systematic risk of Moroccan financial stocks is constant over time and across investment horizons, as the Capital Asset Pricing Model (CAPM) assumes. Using daily data from January 2013 to June 2026, we estimate a time- and frequency-varying beta for two banks (Attijariwafa Bank and Banque Centrale Populaire) and two insurers (Wafa Assurance and AtlantaSanad) against the MASI. The beta is derived from continuous wavelet coherence, signed through the wavelet phase, and compared with static OLS and 60-day rolling betas. A reliability index measures the share of the time-frequency plane over which co-movement is significant against a Monte Carlo red-noise null. A sharp sectoral divide emerges: banks are strongly and positively integrated with the market, especially at longer horizons, whereas insurers display weak, intermittent, and occasionally negative exposure. The reliability index falls from 61.9% for Attijariwafa Bank to 15.4% for AtlantaSanad, and insurer betas become aggressive during low-frequency stress episodes such as COVID-19. A single OLS beta therefore misrepresents systematic risk at most horizons. Investors and regulators in concentrated emerging markets should rely on horizon-specific hedge ratios and treat the diversification benefits of insurers as conditional on calm market conditions.