주택가격 순환주기의 지역 간 관계 분석*

Abstract

The Korean government is implementing regionally segmented housing market regulation. To identify the impact of the regulation on specific regions, research on regional ripple effects in housing market should be preceded. We analyze the ripple effects by using regional housing price business cycles. While most of previous studies in Korea used Hodrick-Prescott(HP) filter to extract cyclical components, we use Christiano-Fitzgerald(CF) filter instead. Cross correlation analysis suggests that the capital area’s housing price business cycle precedes the housing price business cycles of 5 metropolitan cities and other provinces. Granger causality test supports this result, and also suggests that all the regional housing price business cycles of the three regions Granger-cause one another. According to VAR analysis, however, 5 metropolitan cities and other provinces do not cause each other but drift independently. We try to explain these relations among regional housing price business cycles by means of housing demand and supply, GDP and housing mortgage. The result of regression analysis on regional housing price business cycles and the business cycles of other variables shows that there is hierarchical sensitivity to GDP and housing mortgage balance among those regions. Therefore, we narrowly conclude that the ripple effects on the three regions might exist not just because of market expectation, but because of regional differences in sensitivity to market fundamentals.