This study examines the time-scale connectedness between returns on nine African stock markets and commodities markets across energy, agriculture, metals, and beverage. First, we examine multi-scale (short-, medium-, and long-run) wavelet structural relationships between African stocks and commodities using the bivariate wavelet coherence. We establish that commodities and African stock returns co-move across multiple scales and co-integrate in the long run, albeit sparse. Second, we analyze the portfolio performance of the African stock markets with other commodities using wavelet-based diversified and undiversified portfolios in a translation-invariant manner to calculate the scale-specific Sharpe ratios over different sub-periods rather than giving a one-shot look for the entire sample. This enables us to examine how risk-adjusted returns vary across different periods. The results confirm that having a combined portfolio of commodities and equities improves performance over different investment horizons. Specifically, we observe that in non-crisis periods, particularly from 2001-2006 the equally weighted and optimally weighted portfolios show the greatest performances. However, as we enter into the crisis zones such as the Asian crisis of 1997-2000 and the global financial and Eurozone debt crisis the risk-aversion of investors become prominent as the risk-minimizing portfolios record the highest performances. (C) 2020 Elsevier B.V. All rights reserved.