An Empirical Analysis of House Price Bubble: Empirical Results

An Empirical Analysis of House Price Bubble: Empirical ResultsFollowing the 1997 Asian financial crisis, the confidence of Chinese stock market investors was depressed. Because of limited alternatives, investors began to demand and speculate on mortgages. As a result, increasing interest rates would logically have had an impact on housing consumption in the long run. However, demand in the Beijing housing market did not change during our study period. Therefore, given a certain amount of house supply, developers appeared to be more sensitive to changes in the interest rate. When interest rate tracked up, the extent of the upward pressure from house developers would have been greater than the downward pressure from the house buyers. Financial services

The impact of supply cost on Beijing house prices is 0.0013, and is statistically significant at the 5% level in the long run model. This result implies that a unit change (in Chinese RMB) in the cost of supply will affect Beijing house prices by 0.0013%. On average, the Beijing house price is 8117.5 RMB per square meter; when supply cost increases by 1 RMB per square meter, Beijing house prices will increase by 0.11 RMB per square meter. Our findings support the results published by Lv and Guo and Duan. Lv analyzed house price dynamics using income and cost of supply. Guo and Duan modelled house price using cost of supply and the supply-demand relationship. Both studies reported a positive relationship between supply cost and house prices.
Table 3 shows a positive GDP growth coefficient (14.2145) for the short run model, which is significant at the 1% level. This implies that a 1% increase in the speed of GDP growth rate will increase the Beijing house price index by 14.2%. The growth of income will increase housing affordability, which should positively impact the demand for houses. This is illustrated by Yang and Shen, who investigated the Beijing housing market from 1990 to 2005. The authors reported that the income variable is one of the most important determinants of housing affordability in Beijing market. The authors also found that the impact of income on housing affordability is very large, especially for the first-time house buyers. Stone examined the fundamentals of housing affordability in the U.S. market, confirming that the income variable has a significant impact on house prices. The authors reported that gross family income determines how large a property loan the buyer can afford to repay, and that this significantly impacts housing affordability. Similar to the long run model, the inflation variable (45.8414) of the short run model is statistically significant. CPI also measures price factors that impact house prices such as price of raw materials and labor. This approach was utilised by Qiu in examining the relationship between house prices dynamics and CPI in China using an autoregressive distributed lag model, with data covering the period from 2004 to 2010. The author reported a significant positive link between CPI and house prices.
The interest rate coefficient is positive but insignificant. Changes in interest rate hurt the confidence of the real estate market, which is reflected in the demand for houses in the short run (Hu and Guan, 2011). This makes the interest rate statistically insignificant in the housing transaction.
The supply cost coefficient is also insignificant. Generally, building in a residential area includes several procedures such as land purchase, house design, build and sale. This implies the developers require a longer time period to complete the development of a housing tract (from land purchase to the sale of houses). For example, in China, building a housing development for commercial purposes (e.g., subdivisions of houses or multiple apartment buildings) normally takes 10 to 14 months. Therefore, changes in supply costs such as labor and raw material prices will not impact the construction costs of the house currently sold in the short run, especially pre-sale houses, since the developers will not know the actual construction costs. They only know the forecasted construction costs until they sell the house. This forecasted construction costs is predicted using previous construction costs data times the CPI.

Table 3: Estimated Results of the Short Run Model (Equation 3 with quarterly data)

Dependent Variable: HPI
R-squared 0.802037
Adjusted R-squared 0.749247
S.E. of regression 2.346986
Sum squared residual 82.62517
Log likelihood -42.56459
F-statistic 15.19296
Probability(F-statistic) 0.000037
Durbin-Watson stat 1.961852
Coefficient Std. Error t-Statistic Prob.
C 69.89013 8.890478 7.861234 0.0000
LOG(GDPGROWTH) 14.21448 3.020554 4.705919* 0.0003
DLOG(CPI,0,4) 45.84141 18.14815 2.525955* 0.0233
IR 1.465816 1.046576 1.400583 0.1817
SUPPLYCOST 5.78E-05 0.000146 0.395731 0.6979