基于X-12-ARIMA和AR-GARCH模型的房价波动研究

摘要：

以SAS软件为工具,以2007年1月至2015年6月北京市新建住宅价格指数序列为样本,构建时间序列模型进行实证研究.结果表明,基于X-12-ARIMA模型和AR（2）-GARCH（1,1）模型的复合模型是拟合房价的最优模型.房价序列存在明显的季节特征和典型的波动聚集性,X-12季节调整方法和异方差模型显著有效,拟合相对误差不超过0.4%.对房价的短期预测表明,近期内房价仍保持3%~5%的增长态势,且外部因素对房价的影响程度远远大于房价自身的波动冲击力.

Based on SAS software, the paper empirically studies the newly built housing price indices of Beijing from Jan. 2007 to Jun. 2015 by establishing all kinds of time series models. The study results show that the optimal fitting model is the compound model of X 12 ARIMA model and AR（2）-GARCH（1,1） model. There is the clear seasonal fluctuation and sig- nificant clustered volatility in housing price series. The whole effect of carrying on seasonal adjustment by X-12 method is obvi- ous and the heteroskedastic model is particularly valid. The relative error of forecasting is less than 0.4%. The results of the short term forecasting indicate that the housing price will rise in the range of 3%-5% in the near future. Meanwhile, the influ- ence degree of the external factor on the housing price is distinctly stronger than that of the internal impact of the housing price series itself.

作者：

聂淑媛

机构地区：

洛阳师范学院数学科学学院

出处：

《betway官方app 学报：自然科学版》 CAS 北大核心 2016年第4期39-44,共6页

基金：

河南省软科学研究计划项目(152400410152) 河南省科技厅计划项目(162400410122) 河南省高等学校重点科研项目(16B110008)

关键词：

房价 X-12-ARIMA模型 AR-GARCH模型 季节调整

housing price X-12-ARIMA model AR-GARCH model seasonal adjustment

分类号：

O212 [理学—概率论与数理统计]

基于X-12-ARIMA和AR-GARCH模型的房价波动研究.pdf