저자: 손종칠

2000년대 들어 주요 선진국을 비롯한 대다수 나라에서 부동산, 주식
등 자산가격의 급등락 현상이 빈번해 지면서 그 원인과 정책대응 방식
에 대해 논의와 분석이 다양하게 이루어져왔다. 본고는 부동산시장의
불안이 본격 재연되기 시작했던 2001년 이후부터 2008년까지의 기간을
대상으로 주택가격의 변동요인을 거시실물, 주택, 금융부문에서 10개 변
수를 선별하여 각각을 동태적으로 비교 분석하였다.

본고에서는 먼저 분석대상 시계열의 자료 제한을 보완하고 모형의 적
합성과 예측력을 제고하기 위해 Villani(2008)의 베이지언 VAR을 분석
방법으로 원용하였다. 그리고 모형식별은 VAR 변수의 외생성을 보다
객관적으로 도출할 수 있는 DAG(Directed Acyclic Graph) 기법을 활용
하였으며 이를 토대로 충격반응분석, 예측오차에 대한 분산분해 및 역
사적 분해, 조건부 예측 등을 시도하였다.

분석 결과 2000년대 들어 실질 주택가격(국민은행의 전국 주택가격지
수 기준) 변동은 콜금리 변경에 의해서는 불확실하고 미약하게 영향을
받았던 반면 국민소득, 소비지출, 주거용 건설투자, 물가 등 실물경제충
격과 가계대출충격에 의해서는 유의적이면서도 상대적으로 크게 영향을
받은 것으로 나타났다.

이에 따라 앞으로 부동산가격 안정을 위해서는 일관성 있는 거시경제정
책을 통해 실물경제충격을 최소화해 나가는 한편, 부동산부문으로의 자금
유입이 과도하게 일어나지 않도록 시장모니터링 등을 강화해 나갈 필요
가 있다고 판단된다.


Ⅰ. 머리말 ······································································································1
Ⅱ. 주택가격 변동 추이 ·············································································3
Ⅲ. 기존 연구 ································································································6
Ⅳ. 실증 분석 ································································································8
1. 분석모형 ···········································································································8
2. 분석자료 선택과 사전정보 설정·································································9
3. 실증분석 결과·······························································································17
V. 맺음말 ·····································································································33

<참고문헌> ············································································································34

<부록> ·····················································································································38