摘要 本文使用对数周期性幂律(Log Period Power Law, LPPL)模型对房地产市场价格泡沫进行测度,运用空间计量模型对我国房地产市场价格泡沫和空间传染效应进行研究。LPPL模型认为由价格泡沫产生并最终破裂的金融市场与地震系统具有很多相似之处,即金融资产的价格呈周期性变化规律,价格持续上涨到临界状态直至反转。本文采用2010年6月至2017年11月间我国100个城市的房地产市场数据对各城市房地产价格泡沫进行测度和物理/经济空间传染效应研究。研究发现,LPPL模型能够对我国100个城市房地产价格泡沫进行甄别且主要存在两种泡沫状态:正向泡沫(房价持续上升)和反转泡沫(房价整体下降却存在反转点)。各个城市(地区)房地产价格具有较强的空间传染性;存在正向泡沫区域的空间传染性相较反转泡沫区域更为明显,在考虑经济空间测度而不是物理空间测度的情况下,各城市间的空间传染性更强。与现有文献不同,我们发现反转泡沫区域的新房价格指数特别是二手房价格指数的上升对周边城市的房地产价格指数存在强烈的正向推高影响。最后,本文发现城市的房地产调控政策在一定程度上抑制了房价传统影响(比如信贷、新房、二手房价等)因素的推高影响,但各城市房地产价格之间的联动变化特征应该引起监管部门的注意。
Summary:
This study uses the log period power law (LPPL) model to measure price bubbles in real estate markets in China, and the spatial econometric model to study the price bubble contagion effect. This study asks the following questions. Compared with other financial assets, how can price bubbles be measured in the real estate market？ Are price bubbles in the real estate market spatially contagious？ What is the mechanism behind the contagion？ Do macro-control policies effectively prevent real estate price bubbles from expanding？ This study not only quantitatively analyzes the real estate price bubbles in various Chinese cities, but also discovers the relationship between real estate bubbles in different regions. The findings will help local governments to regulate real estate based on local conditions and will help them to develop appropriate policies. According to the LPPL model used in this study, the financial market generated by price bubbles and their eventual collapse is in many ways similar to the seismic system; that is, the price of financial assets changes in a cyclical pattern, and the price continues to rise until a critical state leads to a reversal. The LPPL model is mainly used for seismic research. Johansen, Ledoit, and Sornette (2000) and Zhou and Sornette (2005) are the first to apply LPPL to the analysis of asset price bubble behavior in financial markets. Many studies use the LPPL to predict historical financial crises and bubble and anti-bubble phenomena in Western financial markets. They all find that the LPPL model has the best simulation and prediction results for studies of these phenomena. The LPPL model used in this study is a commonly used and mature model for studying the bubble theory. It uses observed price time series data to detect the formation of price bubbles and their expected collapse point, that is, the end point of the bubble. It focuses on the simulation of the price formation process itself and the prediction of the price reversal point. Unlike the stock market, real estate price bubbles are characterized by medium-to-long-term continuous rises in price formation, and they occur slowly. The LPPL model can better simulate the process of real estate price growth and reversal. Unlike previous studies, this study considers the characteristics of real estate price bubbles in both the upward and downward stages of the bubble. The biggest difference between the two stages lies in the price dynamics before and after the price collapse point. In a positive bubble, the price appears to grow faster than exponentially with accompanying oscillations, and a price collapse point appears at a future point. The reverse bubble stage begins at the price collapse point, after which prices trend downwards. The second innovation of this study is to use the combined cross-sectional physical and economic distances between cities to explore whether real estate price bubbles measured by LPPL have a spatial contagion effect. Finally, this study uses the recent real estate control policies issued by some cities to conduct an event study of the impact of these policies on real estate price bubbles. Therefore, this study examines the behaviors of real estate price bubbles before and after the implementation of real estate control policies in some first-tier cities, and whether the contagiousness of the bubble space is weakened or strengthened by these control policies. This study uses microstructure real estate market data from 100 cities from the June 2010 to November 2017 period. The LPPL model identifies real estate price bubbles in 100 cities in China, and the bubbles have two main states: a positive bubble (housing prices continue to rise) and reverse bubble (housing prices decline overall, but there is a reversal point). In each city (area), real estate prices have a strong spatial contagion. The spatial infectivity of areas with a positive bubble is more obvious than that of areas with a reverse bubble. When economic measures are used to define spaces instead of physical measurements, the spatial infectivity between cities is stronger. Unlike previous studies, this study finds that increases in the new house price index in reverse bubble areas, especially the second-hand house price index, has a strong positive impact on the real estate price index in surrounding cities. Finally, this study finds that the real estate regulation policies in cities to some extent restrain the traditional impact of housing prices (such as credit and new and second-hand housing prices). Purchase restrictions, price restrictions, loan restrictions, and sales restriction all affect the real estate market through different channels. Loan restrictions create an inflection point in the market, whereas sales restrictions can effectively inhibit market investment. However, the link between real estate prices in various cities should be considered by regulators seeking to control real estate price bubbles.
李伦一, 张翔. 中国房地产市场价格泡沫与空间传染效应[J]. 金融研究, 2019, 474(12): 169-186.
LI Lunyi, ZHANG Xiang. Real Estate Price Bubbles and the Spatial Contagion Effect:Evidence from 100 Cities in China. Journal of Financial Research, 2019, 474(12): 169-186.
Asgharian, H., Hess, W. and Liu, L., 2013. “A Spatial Analysis of International Stock Market Linkages”, Journal of Banking & Finance, 37(12), pp.4738~4754.
[12]
Cameron, G., Muellbauer, J. and Murphy, A., 2006. “Bubble trouble-are British House Prices Significantly Overvalued？”, Economic Outlook, 5.8.
[13]
Capozza, D. R., Hendershott, P. H. and Mack, C., 2002. “Determinants of Real House Price Dynamics”, NBER Working Paper Series, 1.1.
[14]
Chan, H.L., Lee, S.K. and Woo, K.Y., 2001. “Detecting Rational Bubbles in the Residential housing Markets of Hong Kong”. Economic Modelling, 18(1), pp.61~73.
[15]
Glaeser, E.L., 2017. “Real Estate Bubbles and Urban Development”. Asian Development Review, 34(2), pp.114~151.
[16]
Johansen A, Ledoit O. and Sornette D., 2000. “Crashes as Critical Points”. International Journal of Theoretical and Applied Finance, April 3(02):219~55.
[17]
Nneji O., Brooks C. and Ward, C., 2015. “Speculative Bubble Spillovers across Regional Housing Markets”, Land Economics, Volume 91, Issue 3.
[18]
Qun Z.., Sornette D. and Zhang, H., 2017. “Anticipating Critical Transitions of Chinese Housing Markets”, Swiss Finance Institute Research Paper, No. 17~18.
[19]
Sornette, D., Woodard, R., Yan, W. and Zhou, W.X., 2013. “Clarifications to Questions and Criticisms on the Johansen-Ledoit-Sornette Financial Bubble Model”, Physica A: Statistical Mechanics and its Applications, 392(19), pp.4417~4428.
[20]
Zhou W. X. and Sornette, D., 2005. “Is there a real-estate Bubble in the US”, Physica A: Statistical Mechanics and Its Applications, 361(1):297~308.
