Purpose This study aims to examine whether Sharia screening criteria can be systematically translated into informative prior distributions for a Bayesian GARCH volatility model and whether such priors improve both parameter estimation and out-of-sample forecasting for Islamic equities. Design/methodology/approach Three Sharia norms (low leverage, prohibition of speculation, real-asset backing) are mapped onto Beta and Inverse-Gamma prior distributions for GARCH parameters. A custom Metropolis–Hastings algorithm, implemented in Mata (the matrix programming language embedded in Stata 19.5), estimates Bayesian GARCH(1,1) and GJR-GARCH(1,1) models using 1,446 daily returns of the HLAL Shariah ETF and S under the QLIKE criterion, Diebold–Mariano tests confirm a statistically significant improvement over the frequentist benchmark and the Model Confidence Set reduces to the Sharia-informed model alone at the 90% confidence level, whereas the informative prior yields no comparable advantage for the conventional benchmark. Robustness analyses, including a GJR-GARCH extension that documents a 29 % lower leverage-effect parameter for Islamic equities, prior sensitivity checks and sub-period partitioning, corroborate the main findings. Research limitations/implications The analysis focuses on US-listed Shariah-compliant ETFs; extension to sukuk, multiasset portfolios and emerging-market Islamic indices would strengthen generalizability. Practical implications Islamic fund managers can adopt the Sharia-to-prior mapping to improve volatility forecasting, particularly during crisis periods. Originality/value To the best of the author’s knowledge, this is the first study to derive Bayesian prior distributions directly from Islamic jurisprudential (FIQH) norms, bridging the gap between Sharia compliance standards and financial econometrics. It provides a replicable methodological template for encoding qualitative regulatory constraints as quantitative priors.
Emin Ahmet Kaplan (Tue,) studied this question.