The housing sector in Korea is currently in transition due to various structural and environmental changes, including the accelerating population aging, the lowering potential GDP growth, and the shifting tenure choice patterns. Given these changes, the current study aims to quantify effects of housing market driven stress scenarios on four specific economic agents of relevancy: 75 listed construction companies, 17 commercial banks, 96 savings banks, and households (or borrowers of residential mortgage loans). We employ particular performance indicators for these agents (operational profit rates for the construction firms, capital ratios and non-performing loan -NPL- rates for the commercial banks, and NPL rates for the saving banks), and estimate a regression model for each of the indicators. In our empirical analyses, we use the quantile regression methods, which yield an interval, rather than point, estimation of the effect of each scenario tested.

As a summary, the results show that: as expected, the stress scenarios employed raise NPL rates of the commercial and saving banks but lower their capital ratios and the operational profit rates of the construction companies; as more micro evidences, the negative effects of the scenarios are more severe for smaller business entities as well as for lower-income and less creditworthy borrower cohorts. In addition, the quantile regression results demonstrate that the model selection bias tends to be large: for example, the NPL ratios of the commercial banks are shown to respond more sensitively with a higher-percentile quantile method, e.g., 2.71 percent increase (in response to a particular scenario) with a 50th percentile regression vs. 4.66 percent with a 95th percentile regression (a 72 percent rise). Based on these and other results, we discuss several implications for performing stress test in general, such as selecting appropriate empirical model (OLS vs. quantile regression), setting threshold for a partical risk indicator (choosing between 99th vs. 95th confidence interval), and establishing international best practices (and performing further research to that end) with regard to these and other analytical issues.

발간사
요 약

제1장 연구의 개요 및 결과 요약(조만⋅송인호)
　제1절목적 및 방향
　제2절실증분석 개요
　제3절결과 요약
　제4절정책 시사점
　참고문헌
　부 록

제2장 스트레스 테스트 시나리오 설정과 거시경제모듈의 적용(송인호⋅김현아)
　제1절스트레스 테스트 시나리오 설정
　제2절거시경제 모듈
　제3절시나리오에 따른 거시모형 반응
　참고문헌

제3장 부동산시장 구조모형을 이용한 예측모형 개발 및 스트레스 테스트(민인식)
　제1절 서 론
　제2절 주택실물-주택금융시장의 구조모형
　제3절 구조모형을 이용한 예측
　제4절 주택시장 예측 및 스트레스 테스트
　제5절 결론 및 정책적 시사점
　참고문헌
　부 록

제4장 부동산시장 침체에 따른 은행, 저축은행, 건설업 스트레스 테스트(민인식⋅최필선)
　제1절 서론
　제2절 데이터 및 스트레스 테스트 모형
　제3절 스트레스 테스트
　제4절 요약 및 시사점
　참고문헌

제5장 부동산시장 가격 충격에 따른 가계건전성 스트레스 테스트(이창무⋅임미화⋅최성호)
　제1절 개별가구 부도확률 결정모형
　제2절 주택가격 변동에 따른 가구소득 효과 추정
　제3절 시나리오별 스트레스 테스트
　제4절 결론 및 정책적 함의
　참고문헌