The Markov switching model on early warning in banking vulnerability: comparison between Indonesia and Malaysia

Purpose This research paper aims to analyze the fluctuating instability of Islamic banking in Indonesia and Malaysia by identifying early warning signs of banking crises, determining the lengthiest crisis durations and evaluating the stability of Islamic banking in both nations. Design/methodology/approach Using the z-score to measure banking stability, this research examines data points spanning from January 2007 to August 2022, using the Markov Switching Vector Autoregression (MS-VAR) method to identify changes in economic circumstances. Findings Findings show Malaysia is more stable, while Indonesia experiences shorter crisis periods. Key factors such as credit, exchange rate, debt, foreign rates and money supply raise the probability of a crisis in both countries. Some variables are not consistently significant, highlighting different stability drivers. The results suggest that tailored risk management is needed in dual banking systems. Research limitations/implications This study is limited by its use of the Z-score as the sole proxy for banking stability, which may not capture all dimensions of financial soundness. Additionally, the analysis focuses only on Indonesia and Malaysia, limiting the generalizability of the findings. Future research should consider alternative or multiple stability indicators and include a broader set of countries for more comprehensive insights. Practical implications The findings highlight the need for stable and proactive regulation in managing Islamic banking risks. Policymakers should strengthen macroprudential oversight, monitor key risk factors and support Islamic banks through capital buffers and risk-sharing instruments to enhance resilience during crises. Originality/value This study contributes to the limited literature comparing Islamic banking stability across countries, particularly in Southeast Asia, using a dynamic econometric approach to capture cyclical fluctuations.