An Empirical Analysis of House Price Bubble: Conclusion

Because of the lagged effect of supply cost and interest rate (current changes in fundamental variables did not capture the real costs of houses currently being sold), developers tend to use CPI to estimate the total costs of housing supply in the short run (based on quarterly data). However, in the long run (based on yearly data), developers already know their actual total costs of the housing supply; therefore, they tend to focus on the interest rate (borrowing costs) and supply cost (construction and operating costs) to make their decision about house supply and house prices. In general, developers make their decisions based on profit, which is defined as house prices minus total costs of their housing supply; therefore, for a given markup, the house supply is dependent on the total costs. As a result, the costs of capital (which depends on interest rate) and construction and operating costs (which depend on supply cost) significantly impact house price in the long run. However, in the short run, due to incomplete information about total house supply costs (such as what happens when most of the houses in a development are sold before they are completely built), housing developers can only use CPI as a proxy of the total costs of the housing supply to estimate their total costs. Therefore, CPI is more effective in explaining house price in the short run. Credit card

Hott and Monnin suggest that to test the existence of a housing bubble one should address the gap between real house price and its fundamental prices. Therefore, in order to provide descriptive evidence of the existence of housing bubbles in the Beijing housing market, we compared the movement between the house price index and the equilibrium house price index, incorporating the economic fundamental variables from our statistical model. Figures 1 and 2 show the movements of the real house price index and the equilibrium house price index from the long run (equation 2) and short run (equation 3) models, respectively.
In the short run model (Figure 2), the dotted line represents the equilibrium house price index in the Beijing housing market, building in the economic fundamental variables of income, inflation, interest rate and construction cost of house supply. The study results show two similar trends. The only significant differences appear in the last three quarters of 2006, in the second half-year of 2007 and 2010. Theoretically, a house price bubble exists when the real house price index is greater than the equilibrium house price index for a relatively extended period (e.g., three consecutive quarters). Therefore, the result of the short run model shows a housing bubble is very likely to have existed in the Beijing housing market from early in 2006 to 2007.
This conclusion from Figure 2 can be supported by the results of the long run model analysis. The large gap between house price and long-run fundamental price suggests the possibility of a house price bubble in the real estate market. In Figure 1, the house price index in 2005, 2006 and 2007 are greater than the equilibrium house price index, which suggests the existence of a bubble in the Beijing housing market. This interpretation of results is consistent with those of Hou. Hou analyzed the relationship among house price, income and house rent. His analysis showed that the price to income ratio (P-I ratio) in the Beijing housing market increased from 7.6 to 17.1 during the 2004 to 2007 period; in the same period, the price-to-rent ratio (P-R ratio) in the Beijing housing market increased from 15 in 2004 to 32 in 2007. The author explained that the P-I ratio that market is more than 50% higher than the average P-I ratio for the developed regions in the eastern provinces of China. In general, the P-R ratio moves between 9 and 18. A ratio above 18 implies a potential price bubble in the housing market. The P-R ratio of Beijing rose over 18 and has remained so since 2005. Therefore, the author concluded the existence of a house price bubble in the Beijing housing market. Similarly, Wu, Gyourko and Deng reported historically high P-I and P-R ratios in Beijing. This was especially so for the P-R ratio, which has experienced a near 70% increase since 2007. Both Chovanec and Xu reported an extremely high P-I ratio in the Chinese housing market in their studies. As a result of our analysis, we can conclude the existence of a house price bubble affecting the Beijing housing market. Assuming that housing markets in very large cities will perform in relatively similar ways, it is quite possible that the results of our analysis will also apply to other large-scale metropolitan cities in China such as Shanghai, Guangzhou and Shenzhen.

Figure-1

Figure 1: Long Run House Price Index (HPI) and Equilibrium HPI in Beijing

Figure-2

Figure 2: Short Run House Price Index (HPI) and Equilibrium HPI in Beijing