Financial Stability Index: A Systematic Literature Review

Abstract

Measuring financial stability became a top priority after the 2008 global crisis, prompting the development of the Financial Stability Index (FSI) as a comprehensive monitoring tool. However, information on the effectiveness of various FSI methodologies is still limited. This systematic review is designed to analyze the components, methodology, and effectiveness of FSI in predicting financial crises. The method used is PRISMA. Articles were obtained from 4 databases, namely: Scopus, Web of Science, ScienceDirect, and Google Scholar. Articles were collected from 2000 to 2025 and 187 articles met the inclusion and exclusion criteria. The results of this study show that (1) there are three main categories of FSI indicators used, namely: CAMELS framework (75.9%), macro-financial ratios (49.7%), and stress-testing variables (35.8%); (2) there were six dominant methodological patterns: panel regression (40.6%), Principal Component Analysis/PCA (47.6%), fuzzy logic (18.2%), machine learning (23.0%), equal weighting (27.8%), and Analytic Hierarchy Process/AHP (16.6%); (3) FSI's predictive ability shows mixed results with Area Under ROC Curve (AUC) ranging from 0.65-0.89 (median 0.76); (4) the main challenges include aggregation issues, weighting controversies, structural heterogeneity between countries, and data quality disparities; Meanwhile, future research recommendations focus on the integration of high-frequency data, hybrid models, non-conventional indicators, and robustness testing. Collaboration between regulators, academics, and practitioners is essential to improve the effectiveness of FSIs.