Validation of Key Factors Affecting Credit Risk for Designing an Early Warning Model in the Iranian Banking System (Case Study: Bank Sepah)

Abstract

Credit risk is one of the most critical challenges facing banking systems, which, under conditions of economic instability and informational constraints, can lead to an increase in non-performing loans and weaken the financial soundness of banks. Despite extensive research on credit risk assessment, a considerable portion of previous studies has relied primarily on quantitative approaches and has paid less attention to the systematic identification of key risk factors based on expert knowledge and practical experience. Accordingly, the present study was conducted with the aim of identifying and building consensus on the key factors influencing credit risk in order to design an early warning model for the Iranian banking system. This study is applied in terms of purpose and qualitative in terms of methodology. The Delphi technique was employed as the primary tool for data collection and analysis. The statistical population consisted of 15 experts in the field of credit risk who were selected through purposive sampling. The panel of experts included senior managers of Bank Sepah, specialists from the Central Bank, faculty members from leading national universities, and chief executive officers of credit rating agencies. The Delphi process was conducted in three consecutive rounds. In the first round, 32 preliminary variables affecting credit risk were identified. In subsequent rounds, using consensus criteria including the within-group agreement index (RWG < 0.15) and the coefficient of variation (CV < 0.25), the variables were gradually refined, ultimately resulting in the extraction of 15 key variables agreed upon by the experts. The findings indicated that the factors influencing credit risk can be classified into three main categories: financial, non-financial, and macroeconomic factors. Experts emphasized the central role of financial indicators alongside customer behavioral factors and macroeconomic conditions in shaping credit risk. The results further demonstrate that reliance solely on financial indicators reduces the ability to identify risks in a timely manner, whereas the application of an expert consensus–based early warning framework can significantly enhance the accuracy of credit decision-making. The conceptual model derived from this study provides an appropriate foundation for designing credit risk early warning systems in banks—particularly Bank Sepah—and can contribute effectively to improving risk management practices and reducing non-performing loans.