【作者】 唐建立；

【导师】 任宏； 张维麟；

【作者基本信息】 重庆大学 ， 管理科学与工程， 2019， 博士

【摘要】 我国房地产自1978年开始发展以来,经历了40多年发展。伴随我国改革开放,经济高速增长到平稳发展,房地产也经历了起步、加速、炒作、调整、调控、反调控、反思与总体整理阶段。在房地产发展过程中,房地产价格总体上呈现不断增长趋势,房地产价格是否真实反映其价值,是否存在泡沫,一直是学界不断讨论和辩证的论题。本文在对房地产泡沫进行理论研究前提下,梳理了房地产泡沫的测度方法,借鉴采用2003年-2013年房地产新的住宅价格指数,基于房地产理性泡沫的前提,采用房地产价格泡沫新的测度方法——PWY和PSY法,重新对我国30个一线二线城市进行房地产泡沫有无的存在性实证测度研究,并根据泡沫发生和破灭时点采用PSY法进行实时测度,为房地产政策制定者提供决策依据。在理论分析房地产泡沫影响因素的基础上,基于动态短面板模型,分析了我国房地产泡沫及其影响因素的联动关系,同时对我国不同区域进行泡沫城市间的空间传递进行研究,采用面板动态空间计量模型,对我国不同城市房地产泡沫空间效应进行了实证分析。论文利用房地产住宅价格指数新数据,以及相应的租金价格指数数据,采用PWY和PSY方法,对我国30个城市2003年1月到2013年12月的变量数据,进行泡沫实证研究。其实证结果表明,我国30个城市在样本研究期间,大部分城市存在不同程度的房地产泡沫,如果在5%显著水平下,只有哈尔滨和沈阳2个城市通过检验,没有发现具有泡沫;如果在10%显著水平下,仅有哈尔滨1个城市被拒绝,没有房地产泡沫,其余城市均存在一定程度的房地产泡沫,并且不同城市房地产泡沫也存在一定差异。通过区域性分析发现,我国中部和西部区域房地产有泡沫城市比例程度明显高于东部区域比例。同时,通过PSY法对30个城市的住宅房地产泡沫进行了实时测度,这为政府进行削峰平谷式的宏观调控提供政策决策依据,也为政府实施预警机制提供了有效的检测方式。论文在基于房地产泡沫测度结果的基础上,采用年度内月份出现有泡沫的比例来计量当年房地产泡沫大小的程度,基于多个房地产泡沫因素,30个城市的相关数据,构建了动态短面板模型进行实证研究分析,结果表明人均储蓄、人均土地出让金、土地出让金楼面地价、人均住宅房地产投资等因素均显著正向影响房地产泡沫,而人均住宅竣工面积、人均GDP、人均可支配收入等因素却反向显著影响房地产泡沫,但人口数量、人均国内贷款、通货膨胀、人均销售土地规划建筑面积等因素对房地产泡沫的形成影响程度概率偏小。论文在考虑各城市空间相邻权重的基础上,分析了房地产泡沫的空间效应,结果发现,各个城市间的房地产泡沫并非是随机独立的,受到其他地区房地产泡沫以及其他经济等相关因素行为的影响,我国30个城市的房地产泡沫之间存在显著的空间相关性。通过面板空间计量模型的选择,结果发现,构建面板空间德宾模型是最适合我国房地产泡沫额空间效应分析,并运用该模型进行实证分析。结果表明,我国房地产泡沫空间关联效应中,泡沫滞后项、泡沫空间交互项、人均收入、通货膨胀、人均房地产投资、土地出让金楼面地价和土地出让金楼面地价空间交互项对房地产泡沫均有显著影响,同时,采用面板空间德宾模型计算出相应变量的直接效应、空间溢出效应和总效应,发现各自变量所带动的空间溢出效应占总增长效应比例在13%-59%之间,进一步论证了各影响因素所带来的空间溢出效应对我国房地产泡沫产生的重要贡献,其中空间溢出效应最大的是土地出让金楼面地价,其空间溢出效应比例占到59%,这给政府对房地产进行宏观调控给出一定方向。本文采用新的房地产价格数据,新的房地产泡沫测度方法,实证研究测度了我国房地产泡沫,构建了动态短面板模型分析了房地产泡沫的影响因素,利用空间计量模型,论证了房地产泡沫的空间效应。本文进行房地产泡沫测度研究时,没有考虑房地产泡沫破裂对房地产和经济的后续影响;在研究房地产价格泡沫的地理空间传导机制时,没有探讨研究不同集群区域之间,如东部、中部和西部之间的空间效应及传动关系;由于受限于统计数据的原因,本文研究范围仅限于我国30个一线、二线大中城市,相应面板分析和空间效应分析也止步于前几年;在研究对象上,仅集中在住宅房地产市场,未对商业、写字楼等非住宅房地产泡沫进行研究。这些需要在后续研究中不断完善和深化。更多还原

【Abstract】 China’s real estate has experienced about 40 years of development since 1978.With China’s reform and opening up,real estate has undergone the stages of starting,acceleration,hype,adjustment,regulation,counter-regulation,reflection and overall collation.Generally,the real estate price shows a growing trend.However,whether the real estate price truly reflects its value and whether there is a bubble have always been topics of constant discussion in the academic circle.Established against such background,this paper,employing new real estate price indices in the sample period 2003-2013 and utilizing new real estate bubble measurement methods---PWY and PSY methods,based on the premise of rational real estate bubble,has conducted an empirical study on the existence of real estate bubble in 30 first-tier and second-tier cities in China.The study has provided real-time measurement of real estate bubble at times of bubble occurrence and burst point.Based on the theoretical analysis of the influence factors related to real estate bubble,adopting dynamic short panel model,the present study has also analyzed the relationship between real estate bubble and its influencing factors.Finally,using a model of panel dynamic spatial econometric,it has made an analysis of space effect of the real estate bubble between different cities in China.Employing PWY and PSY methods,this paper has conducted an empirical study on the real estate bubble of 30 Chinese cities during the sample period from January 2003 to December 2013 based on the new data on residential real estate price index constructed and data on the corresponding rent price index.The results show that most of the 30 cities have real estate bubbles of varying degrees during the sample period.If the significant level is set at 5%,only two cities,namely Harbin and Shenyang,are found to have no real estate bubbles.If the significance level is set at 10%,only one city,Harbin,is rejected and there is no real estate bubble,while all the other 29 cities have bubbles to some extent,and the bubbles in different cities are different.Further regional analysis indicates that the proportion of cities with real estate bubbles in central and western China is significantly higher than that of the cities in eastern China.Meanwhile,PSY method is also used to measure the residential real estate bubble in 30 cities in real time,which provides a basis for the government to carry out macro-control policy of cutting the peak and leveling the valley,and provides an effective way for the government to implement early warning mechanism.Besides measuring the real estate bubble,taking the annual proportion of months in which real estate bubble occurs as the bubble size,based on the multiple factors influencing real estate bubble and relevant data on 30 cities,this paper has constructed a short dynamic panel model to analyze factors influencing real estate bubble.The results indicate that per capita saving,land leasing,per capital leasing floor price,per capita residential real estate investment have a significant and positive impact on the real estate bubble,while per capita completed housing area,per capita GDP and per capita disposable income have a significant and negative impact on the real estate bubble and factors including population size,per capita domestic loans,inflation and per capita planned floor area of sold land have little influence on the formation of the real estate bubble.In addition,taking into account the urban space adjacent weights,the paper has analyzed the effect of the real estate bubble.The results suggest that real estate bubbles among the cities are not random,or independent of the real estate bubbles of other areas.Instead,they are influenced by the bubbles of other cities and factors related to economic behavior.There exists significant spatial correlation between real estate bubbles in the 30 cities.Through the selection of the panel spatial econometric model,it is found that the construction of the panel space durbin model is the most appropriate.Results of empirical analysis suggest that among the spatial correlation effects,the real estate bubble foam lag,the spatial interaction terms,the per capita income,the price level,the per capita real estate investment,per capita land leasing and land leasing spatial interaction have a significant influence on the real estate bubble.Meanwhile,panel space durbin model was used to calculate the corresponding variables of direct effect,spatial spillover effect and total effect.It is found that spatial spillover effect brought about by each variable accounts for between 13%-59% of total growth effect,further confirming the impact of spatial spillover effect brought by various factors on the real estate bubble in China.Among these factors,the land leasing price of the floor has the largest spatial spillover effect,constituting 59% of the spatial spillover effect,which provides a direction for the government to carry out macro-control on the real estate.Based on the new real estate price index data and the new real estate bubble measurement methods,this paper has empirically measured real estate bubble in China,analyzed its influencing factors,constructed dynamic short panel model and the spatial econometric model,and demonstrated the spatial effect of the real estate bubble.However,this paper is not without limitation.When studying real estate bubble,it did not consider the subsequent effects of the real estate bubble on the real estate and the economy.Also when studying the spatial transmission mechanism of the real estate price bubble,it did not explore the spatial effect and transmission relationship between different clusters,such as the eastern,central and western regions.Due to the lack of data,the corresponding panel analysis and spatial effect analysis cannot extend to the present time.Because of the lack of data,non-residential real estate bubble is not investigated.Therefore,these need to be examined in future research.更多还原