서울 오피스 보유기간에 관한 연구

Abstract

We analyze data representing 73 total transactions in the office building market in Seoul from 1999 to 2008 using a multinomial logit model with panel data. We first estimate three time-varying(t),case-dependent(i) explanatory variables—transaction_pricet,i, cap_ratet,i, and chosei_pricet,i—using market data representing the individual value-weighted market price, the market cap rate, and the market chonsei price per annum, and per 3.3㎡. We found that if the office building sales market is hot, that is, if it features lower funding rates,lower market vacancy rates, lower market cap rates, and higher transaction prices, the selling hazard tends to increase in a statistically significant sense. Second, if transaction costs—measured by gross floor area size, building height, and the reciprocal of the common area ratio—increase,the selling hazard decreases accordingly. Third, older buildings and lower-rated buildings exhibit a lower selling hazard. Based on these results, we conjecture that the Seoul office market is not information efficient because those two characteristics—age and low ratings—are not generally accepted as selling hazard function components. Instead, they affect transaction prices.