공간자기상관을 고려한 수도권 지식산업센터 매매가 및 임대료 모형 구축에 관한 연구

Abstract

The purpose of this paper is to test spatial autocorrelation of sale prices and rents of factories in the knowledge industrial centers(KICs) and establish spatial regression models. First, we collect asking price and rent data of KICs in Seoul metropolitan area from the internet site of the real estate consulting firm, R114. Second, we establish hedonic price equations for total sale price and rent data. By ordinary least square estimation(OLS), we find some statistically significant independent variables like building floor area, age, industrial zone, factory location floor, factory facility area, the adjacency of crossroad, the distance of Seoul City Hall, the distance of subway station and sub-markets. Sub-markets are classified to Guro-Geumcheon, Seongdong, Yeongdeungpo, Seongnam and Gyeonggi southern area. Third, we find strong spatial autocorrelation for median data of prices and rents by Moran’s I and the LISA of Ansellin. On the basis of this result, we establish hedonic equations for median data and examine spatial autocorrelation for errors of OLS. We finally find spatial error model(SEM) the best specific model by Robust LM tests. SEM is turned out to be more specific than OLS model by comparing AC, SIC, Log-likelihood values.