중국 주택가격의 결정요인에 관한 실증분석

Abstract

The reform and opening-up policy was established 40 years ago and China has maintained a rapid economic growth rate since. However, its private housing market did not exist until 1998. On July 3, 1998, the State Council created the Notice on Further Deepening Urban Housing System Reform and Speeding up Housing Construction Act. To remove barriers to the commercial housing market, the reform abolished the housing distribution system. And, in the 20 years since the establishment of the private housing market, it has experienced rapid expansion. Since 2000, the investment in real estate development has continued to rise, from 5.988 billion yuan to 15.6 trillion yuan in 2017, a 25.02-fold increase. In the same period, the investment in real estate development increased from 5.98 percent to 19 percent of the GDP.

In the early years of China’s policy makers, strong policy support enabled the supply of state-owned land to provide capital. This vigorous policy support served to capitalize on the state-owned land supply and provided credit both for developers and consumers, leading to a tremendous expansion of the supply of housing and generating a real estate boom that persisted for more than two decades. However, as China's real estate market has matured, with its enormous size and abundant credit, inventories, and market activity, this style of short-term, heavy-handed intervention can disrupt effective market signals and result in distortions in credit and resource allocations that weigh heavily on the prospects for long term growth. Also, with such a rapid development of the real estate market, many unique situations arise (e.g. “booming tier-1 cities” vs “ghost cities”).

Facing these unique situations, China’s policy makers at many levels have been actively working to improve the functioning of housing markets across a wide variety of geographies and economic circumstances(e.g. Six real estate market regulation measures (2006), Four real estate market regulation measures(2009), Eleven real estate market regulation measure(2010), etc.). Despite the efforts carried out by the government, the housing price level in China relative to household income is still very high reported to the rest of the world and, in general, China's economic structure is vulnerable to the impact on housing prices. On the other hand, the dramatic fluctuations in housing prices in China's major cities have drawn global attention as the country's economy becomes more important and connected to the world economy.

Therefore, to take into consideration the importance of real estate in China's economy and the impact on housing prices on people's lives, we have created three essays on China's housing prices. The first essay conducts an empirical study on China's housing markets, attempting to analyze housing price spillover and the time-varying properties of housing price spillover. The second essay is an analysis of long and short term influencing factors on China's housing prices considering market size and market characteristics. And the third essay is about the asymmetric Long and Short-run Connectedness between China's housing prices and stock price.

In the first essay, all of the variables were selected over the period from July 2005 to October 2017 . We chose the new housing prices index from the 70 cities housing sales price index released by the National Bureau of Statistics to represent housing prices. The generalized VAR and the corresponding generalized spillover index methodology recently introduced by Diebold and Yilmaz(2012) is applied in the current study. Through this analysis, several noteworthy results are found. First, total spillover indexes exceed 40% in both models. Second, the spillover from others has a significant negative relationship with the urban development level, while the spillover to others has a significant positive relationship with the urban development level. Spillovers from other countries are particularly prominent in Beijing. Third, in the time-varying dynamic analyses, the empirical results show that the influence of the first-tier cities reveals a steady declineand the impact of second- and third-tier cities displays an overall upward trend. Within the first-tier cities, the independence of Guangzhou has weakened and the independence of Shenzhen has strengthened. In Shanghai, expansion of influence on the other regions has been verified.

In the second essay, the commonly used framework for modeling housing prices is the life-cycle model (see Meen (2001, 2002), Muellbauer and Murphy (1997, 2008)). All of the variables selected are from January 2007 to October 2017 due to the lack of monthly data on household loans. We chose the new housing prices index from the 70 cities housing sales price index released by the National Bureau of Statistics to represent housing prices. The interest rate is the benchmark lending rate for one to three years and industrial added value was used to indicate this variable due to the lack of monthly data on household income. Completed real estate (housing market) area represents the housing supply. The Shanghai composite index is employed for stock price. In order to study the existence of a long-run relationship of the theoretical model, we performed Johansen consideration and the ARDL Bounds test. Additionally, DOLS and FMOLS analysis was used to determine long-run vectors. To analyze the short run influential factors, generalized impulse response and generalized variance decomposition techniques were employed. Through this analysis, several noteworthy results were found. First, based on Johansen consideration tests, Long-run relationships have been established on all models. But, on the ARDL Bounds test, only first-tier cities have long relationships. Second, for the long-term vector, the positive effects of income and household loans and the negative effects of interest rate were mainly identified in the first-tier cities. The stock price and housing supply were not identified. For second-tier cities, the positive effects of income, household loans, housing supply and adverse effects of stock prices were mainly identified. Interest rate was not identified. For third-tier cities, only stock price is negative. Third, we found that the short run was almost identical as in the long run, except for stock prices and interest rates. Interest rate had a positive influence in second - and third-tier cities. Stock price had a positive impact in first-tier cities. The consequences of the generalized variance decomposition support the generalized impulse response.

In the third essay, all of the variables selected are the same as the second essay and the nonlinear autoregressive distributed lag (henceforth, NARDL) methodology recently introduced by Shin et al. (2014) was applied. Generally, considering the sensitivity of stock price as in our preceding papers, we have divided the whole period into two periods in order to understand the relationship between housing prices and stock price. The entire period was from January 2007 to October 2017. The first period was taken from July 2005 to April 2012 and the second period was from May 2012 through October 2017. Monthly data were used to analyze short- and long-run relationships between the housing market and the stock market. However, due to the market size and market characteristics in the China's real estate market, it was hard to establish second- and third-tier city relationships. Therefore, this essay focuses on first-tier city relationships. Through this analysis, several noteworthy results were found. First, in the entire period and the first period, the long-run rise in stock price had a positive impact on the housing price of first-tier cities, while the long-run decline in stock price had a negative impact on the housing price of first-tier cities. The long-run positive effect is greater than the long-run negative effect. The short-run rise in stock price had a positive impact on the housing price in first-tier cities, while the short-run decline in stock price had a very weak positive impact on the housing price in first-tier cities. In the second period, although there is a long-run relationship between stock price and housing price in the first-tier cities, the long-run change of stock had no impact on the first-tier cities. The short-run change of stock had a relatively weak positive impact on first-tier cities. Second, with second- and third-tier cities, the long-run symmetry and asymmetry of stock prices had no impact on housing prices. Stock prices had a relatively weak impact on second- and third-tier city house prices. Overall, the impact of second- and third-tier cities displays an overall upward short run.

This paper discussing the influence of Chinese housing prices can provide the Chinese Government with more targeted policy recommendations for the regulation of China's housing market.