비전통적 부도예측 모형: 비 재무적 요인과 내부 성과조정 요인

By utilizing data from the Korean Statistical Information Service (KOSIS) for the years 2020 to 2021 and employing the Propensity Score Matching (PSM) method, the study selected defaulted and non-defaulted firms with similar financial conditions and conducted predictive analyses to enhance the reliability of the research. This approach opens up the possibility of constructing more realistic prediction models that consider various non-financial factors, rather than relying solely on financial data. The study results showed that prediction models incorporating non-traditional financial factors had higher accuracy when financial conditions were similar, playing an important role in overcoming the limitations of existing models. The findings are as follows. First, models incorporating non-traditional corporate information significantly improved the performance of default predictions. Second, investments via related-party transactions, both domestic and international, had a positive impact on corporate defaults. Thus, firms’ internal adjustment efforts affect default risks significantly. Specifically, firms with excessive foreign investments were more vulnerable to economic shocks, thereby increasing the likelihood of default. Third, the effect of non-cash incentive systems on corporate defaults was not significant. Based on these findings, the incorporation of a variety of non-traditional and internal adjustment factors to differentiate between groups with similar financial conditions resulted in improved prediction accuracy in the models.