Portfolio optimisation of Shariah-compliant samples in the USA

Purpose This paper aims to investigate the dynamic portfolio optimisation performance of numerous samples of Shariah-compliant firms in the USA vis-à-vis the overall conventional sample. Design/methodology/approach This paper constructs efficient frontiers and subsequently the capital market line using the ovport set of commands in STATA. From the capital market line, the tangent portfolio is found, and the Sharpe ratio of the tangent portfolio is the primary measurement of the dynamic portfolio optimisation performance of the samples of Shariah-compliant samples in this study. Findings This paper finds that the overall conventional sample will outperform the Shariah-compliant samples in most cases. However, there exists a consistent trend whereby the performance of the overall conventional sample will converge towards the performance of the Shariah-compliant samples (and even be lower at times), as the market approaches a looming crisis suggesting that the Shariah-compliant samples do not experience significant deteriorations in their performance as compared to the conventional sample and that they provide stability during such times. Research limitations/implications This paper assumes no transaction costs, illiquidity, bid-ask spread and non-compliant revenue purification all of which may negatively affect portfolio performance. Practical implications The findings of this paper suggest that Shariah-compliant samples should be included in portfolios during times of crisis because they are less affected by market-wide volatility. Social implications The stability of Shariah-compliant samples reflects the conservativity of the contemporary Shariah stock screening methodologies and the Shariah itself. Originality/value Portfolio optimisation studies on Shariah-compliant samples are usually static in nature and are conducted in selected Muslim countries. This paper studies the dynamic portfolio optimisation in the USA where a liquid Islamic capital market is non-existent.