Latent Variable Estimation and Frontier Efficiency Analysis: A Comprehensive SEM-SFA Framework Validating CAMEL Model Efficacy Across Colombian Financial Cooperatives

Abstract

This study presents a comprehensive analysis of financial health and managerial efficiency across 153 Colombian financial cooperatives using integrated latent variable estimation through Confirmatory Factor Analysis (CFA) and Structural Equation Modeling (SEM), coupled with Stochastic Frontier Analysis (SFA). The CAMEL model (Capital Adequacy, Asset Quality, Management, Earnings, Liquidity) serves as the conceptual framework for identifying and quantifying unobservable dimensions of institutional performance. The CFA model explains 89.2% of variability with RMSEA=0.108, validating the theoretical construct of the CAMEL framework. SFA decomposition reveals that 85.2% of efficiency variance is attributable to technical inefficiency rather than random variation, indicating that performance differentials are primarily driven by controllable management factors. Cluster analysis identifies five efficiency tiers, with 12% of institutions achieving very high efficiency (EFI: 90.15%) compared to 5% facing severe distress (EFI: 68.35%). Liquidity emerges as the dominant efficiency predictor (β=1.9247; p<0.001), with coefficient nine times greater than earnings, suggesting that treasury management constitutes the primary lever for institutional performance improvement. These findings validate the CAMEL model's efficacy in emerging market financial systems while providing actionable intelligence for regulatory supervision and institutional management optimization.