Network DEA를 적용한 부동산PF 위험관리와 저축은행 경영성과 분석 : 위험의 편중과 분포 특징

This study applies the Network Data Envelopment Analysis(Network DEA) model to quantitatively assess the management risk of 79 Korean saving-banks over the period from 2018Q1 to 2024Q3. Specifically, the overall management risk(Network_Risk) is decomposed into economic risk (Econ_Risk) and risk management risk(Manage_Risk), with the latter further divided into general loan risk(Risk_GL) and PF loan risk(Risk_PF) for empirical investigation. The analysis reveals that Network_Risk remained relatively stable until 2022Q3, supported by increased loan demand under low interest rates, eased repayment burdens, and improvements in asset quality. However, following the onset of interest rate hikes and a real estate market downturn in late 2022, a marked upward trend in overall risk was observed. Notably, this increase was primarily driven by Manage_Risk, rather than Econ_Risk, indicating that vulnerabilities were more closely associated with loan management practices than with cost efficiency. Risk_PF consistently recorded elevated levels throughout the study period, in direct proportion to exposure volume. Since 2023, concerns over potential defaults in large-scale PF investments, particularly in the Seoul metropolitan area, have intensified. Furthermore, significant heterogeneity was found in the risk management capabilities across saving-banks, with most banks—except for a select few—exhibiting limited capacity to effectively manage PF risks. Importantly, medium-sized saving-banks demonstrated the highest levels of Risk_PF, suggesting an urgent need for enhanced and systematic risk management frameworks tailored to their operational scale and exposure profile.