Credit Risk Modeling: Evidence from the EMV Model Applied to the Autonomous Region of Madeira (2019–2025)
Abstract
This study aims to analyze the evolution and determinants of credit risk in the Autonomous Region of Madeira (RAM) between 2019 and 2025, applying the EMV (Exogenous–Maturity– Vintage) model and robust econometric techniques. The research is based on quarterly data on loans, default rates, and deposits from financial institutions in the region, and compares them with the Portuguese national context. The EMV model allows us to decompose the default rate into three main components: (i) exogenous effects associated with the economic cycle, (ii) maturity effects related to the aging of credit portfolios, and (iii) generation effects (vintage), which capture differences between lending periods. In addition, RLM-Huber models (robust linear regression), quantile regressions, and the Theil–Sen estimator are applied to estimate trends and assess the persistence of defaults. The results reveal that the nonperforming loan ratio exhibits strong temporal inertia, a positive impact of credit growth on delinquency, and a negative correlation between deposits and credit risk, suggesting that savings play a stabilizing role. The conclusion is that the RAM maintains credit quality standards similar to those of the national context, but is more sensitive to economic shocks due to its regional production structure and high dependence on the tertiary sector.
Keywords
References
More Articles from IIARD INTERNATIONAL JOURNAL OF BANKING AND FINANCE RESEARCH
Author: Stephanie Nguhemen Gbande, Mike T. Soomiyol, Timothy Tyona, Gaius Msendoo Asombo
Author: Nguyen Mai Phuong
Author: 1Nwangwu, Okwudiri Iheanyi, 2 Pepple, Grace Jamie PhD
Author: Peter Ngbede Ajam, Abiodun Edward Adelegan, Benson Emmanuel
Author: i, Azolike Nkiru Nkechi, ii, Okwor Emmanuel Ejimnkonye, iii, Eneaniofu Daniel Mmaduakonam
