How Savings－Bank Dependence Interacts with Asset Levels to Shape Consumer Loan Risk

[Purpose] The borrower’s savings bank dependence can be measured as the share of total loan volume that comes from savings banks, indicating how heavily a borrower relies on non－bank financial institutions. This study aims to empirically analyze the determinants of consumer debt workout applications among unsecured loan borrowers, with particular attention to the interaction effects of savings bank dependence, asset levels, homeownership, and age on credit risk. [Methodology] Utilizing data from a lending institution in South Korea, borrowers’ savings bank dependence is classified into four categories：non－users, low, medium, and high dependence. We use multivariate logistic regression with interaction terms for asset level, homeownership, and age. [Findings] While higher savings bank dependence is generally associated with a significantly increased workout risk, the relationship is not simply linear but varies across groups. The risk rises at low dependence, and escalates nonlinearly in the medium and high dependence groups. Asset level and homeownership generally reduce risk, with the buffering effect of assets most pronounced among high－dependence borrowers. Among age groups, older high－dependence borrowers exhibited the highest risk, whereas the effect of high dependence among younger borrowers is relatively more moderate. [Implications] This study demonstrates that the risks inherent in savings bank－oriented lending structures are not uniform, but differ nonlinearly depending on key borrower characteristics and their interactions. These results underscore the need for segmented credit risk assessment and targeted policy interventions.