Robust Estimators in Portfolio Optimisation: Evidence from the Zimbabwe Stock Exchange

Purpose: This study evaluates robust parametric estimates as model inputs for portfolio optimisation relative to the classic sample mean and covariance on the Zimbabwe Stock Exchange, one of the most volatile financial markets in the world. Design/Methodology/Approach: The study uses a quantitative, empirical research design on the Zimbabwe Stock Exchange, one of the most volatile emerging markets in the world. Two portfolios, the traditional Markowitz using sample mean, covariance and the robust portfolio using robust, shrinkage and hybrid estimators are computed and compared their performance over time. Findings: Our findings demonstrate that the combination of robust location and scatter model inputs produces the best balance of return, stable, and robust portfolio with high breakdown points, outperforming the traditional Markowitz portfolio model. Implications/Originality/Value: The Zimbabwean financial markets are highly uncertain and under researched markets, This study evaluate robust estimators directly in such markets, relative to the traditional portfolio optimisation techniques making the comparison both timely and relevant.  It contributes to the literature on portfolio construction under high uncertainty, offering insights for investors and regulators that, it is possible to constru