로짓모형과 오차수정모형을 이용한 아파트가격 상승확률 예측

Predicting the Probability of Apartment Price Increase by Using the Logit & Error Correction Model

This paper predicts the probability of apartment price increase in Seoul by using a Logit & Error Correction Model. In order to analyze the Error Correction Model, market equilibrium prices are estimated by using a reduced-form equation arising from the equilibrium condition of demand and supply. The results suggest that the estimated market equilibrium prices equation fits the data very well and variables in the equation are statistically significant at the 1% level. For analyzing the short-run dynamic features of apartment prices, the Error Correction Model is built and estimated. The Error Correction Model uses the estimates of the market equilibrium prices from the first-stage. The empirical results show that apartment prices exhibit a strong serial correlation, 0.355 and the rate of mean reversion is 0.251. Those independent variables in the Error Correction Model are, then, used for independent variables in the Logit model to predict the probability of price increase. Dependent variable(Y) in the Logit Model is apartment price changes. If the price increse, the value of Y is 1 and if decrease, Y is 0.