导　　师： 任兆璋

学科专业： 020204

授予学位： 硕士

作　　者： ;

机构地区： 华南理工大学

摘　　要： 风险是投资与生俱来的特性,自2007年次贷危机以来,它再一次称为人们关注的焦点。作为一种比较成熟的风险管理工具,VaR已经在实践中得到了广泛的应用,但目前我国很少有人将之用于房地产价格波动风险的评估中。这一方面是由于我国房地产市场还处于初生的高速发展期,还不够成熟;另一方面房地产价格自身的一些特性也对其风险价值评估提出了挑战。然而,房地产作为一种逐渐成熟的投资品,其价格波动风险将会得到越来越多的重视,相应的量化评估也亟待解决,本文则致力于解决这一问题。 本文研究了我国房价短期/(月度/)波动的影响因素,采用的方法主要包括向量自回归模型,协整检验,格兰杰因果检验,动态参数预期等。发现我国房地产市场的波动主要与预期、政策冲击以及利率有关。在此基础上,又以深圳为例探讨如何利用VaR评估其价格波动风险。着重解决了一、二手房价格波动不一致问题,日交易均价的“伪波动”问题,以及房屋月度环比售价指数的“尖峰厚尾左偏”分布的拟合问题。其中采用了蒙特卡罗模拟、混合正态参数极大似然估计的EM法、基于Anderson-Darling检验的经验分布拟合法、以及GARCH-N和GARCH-GED模型。最后,采用各方法估计出的VaR在可合理解释差异范围内,通过月度数据和日交易数据计算出的VaR分别偏于乐观和保守,主张在实际应用当中应该结合各种方法的结果综合评估。至此,本文基本解决了房地产投资中价格波动风险的评估问题,可以供投资者或其他相关人士在进行房地产价格波动风险评估时参考。 当然,本文所做的只是基础性工作,在实际应用中还有待进一步改进和优化,例如对伪波动更好地剔除,对收益序列更优地拟合,实际中的交易摩擦和损失问题等。相信随着未来数据序列长度的延伸,数据质量的提高,研究方法的改进,研究的逐步深入拓宽,房地产投资风险必将得到更好的诠释。 Risks is a inherent characteristic of investment. Since the 2007 subprime crisis, it again called people’s attention. As a risk management tool, Value at Risk /(VaR/) has been in practice applied for a wide range. But very few people in China apply it in the risk assessment of price fluctuateation for real estate. This is because the real estate market is still in nascent high speed development period and is not mature. The other hand, some of real estate’s own characteristics challenges the application of VaR. However, with real estate market’s develop- ment and mature, its price volatility will attract more and more attention, corresponding to the need of the quantitative risk assessment.? This paper is to address this issue. This paper studies the factors in real estate’s short-term /(monthly/) price fluctuations, the adopted methods included: Vector Auto-regression Model, Granger Causality test, Co-integration theory, Extrapolated Expectation based on dynamic parameter, etc. Studies reveal that fluctuations in the real estate market are related to expectation, policy impacts and the interest. Then, on this basis, take Shenzhen as a case study, discussing how to using VaR to assess the risk of price fluctuation. This paper focus on solving such problem as below: the inconsistencies first//second-hand house’s price volatility, the“pseudo-?fluctuation”problem in day trading average price, and the distribution fitting of monthly housing price index, which has“sharp peak,“heavy tail and? negative skewness”. The main methods include Monte Carlo simulation, ML estimation of parameters in mixed-normal distribution based on the EM algorithm, fitting for the empirical distribution based on Anderson-Darling test, and the GARCH-N and GARCH-GED model. Finally, the VaR calculated by those methods are satisfactory, and the fine differences between can be reasonably explained. Basely the assessment of the real estate investment in the risk of price volatility is successful. This paper just raised a basic program and more work is needed to perfect it. I believe that in the future, with the extension of the length of data series, the improvement of data quality and research methods, the quantitative assessment of the real estate investment risk will be better interpreted.

关 键 词： 房地产 风险价值 分布拟合

分 类 号： [F293.3]

领　　域： [经济管理]