当今中国和世界正处于百年未有之大变局,维护国家的经济稳定性至关重要.众所周知,严重的通货膨胀是经济不稳定的一个重要因素.因此,建模和预测通货膨胀率成为了亟待解决的问题.在本文中,我们研究了包括中国和美国在内的世界四个主要国...当今中国和世界正处于百年未有之大变局,维护国家的经济稳定性至关重要.众所周知,严重的通货膨胀是经济不稳定的一个重要因素.因此,建模和预测通货膨胀率成为了亟待解决的问题.在本文中,我们研究了包括中国和美国在内的世界四个主要国家近十年的消费者价格指数(CPI)通胀率,提出含结构性断点波动率与时变参数均值随机波动模型(stochastic volatility in mean model with time-varying parameters and structural breaks in the volatility, SB-TVP-SVM),并给出了相应的贝叶斯估计框架.在以往的大多数研究中,研究者们往往忽略了非平稳特征同时存在于CPI通胀率的条件均值和波动率序列中的可能性.通过引入不可观测的结构性断点, SB-TVP-SVM解决了这一问题,从而得到相比于既有方法更高的序列预测精度.我们模型估计出的结构性断点与过去十年来最大的全球事件高度相关,例如新冠病毒疫情以及俄乌地区冲突.展开更多
This paper proposes an empirical likelihood based diagnostic technique for heteroscedasticity for semiparametric varying-coefficient partially linear models with missing responses. Firstly, the authors complement the ...This paper proposes an empirical likelihood based diagnostic technique for heteroscedasticity for semiparametric varying-coefficient partially linear models with missing responses. Firstly, the authors complement the missing response variables by regression method. Then, the empirical likelihood method is introduced to study the heteroscedasticity of the semiparametric varying-coefficient partially linear models with complete-case data. Finally, the authors obtain the finite sample property by numerical simulation.展开更多
文摘当今中国和世界正处于百年未有之大变局,维护国家的经济稳定性至关重要.众所周知,严重的通货膨胀是经济不稳定的一个重要因素.因此,建模和预测通货膨胀率成为了亟待解决的问题.在本文中,我们研究了包括中国和美国在内的世界四个主要国家近十年的消费者价格指数(CPI)通胀率,提出含结构性断点波动率与时变参数均值随机波动模型(stochastic volatility in mean model with time-varying parameters and structural breaks in the volatility, SB-TVP-SVM),并给出了相应的贝叶斯估计框架.在以往的大多数研究中,研究者们往往忽略了非平稳特征同时存在于CPI通胀率的条件均值和波动率序列中的可能性.通过引入不可观测的结构性断点, SB-TVP-SVM解决了这一问题,从而得到相比于既有方法更高的序列预测精度.我们模型估计出的结构性断点与过去十年来最大的全球事件高度相关,例如新冠病毒疫情以及俄乌地区冲突.
基金supported by the National Natural Science Foundation of China under Grant Nos. 11471060 and 11871124the Key Project of Statistical Science of China under Grant No. 2017LZ27。
文摘This paper proposes an empirical likelihood based diagnostic technique for heteroscedasticity for semiparametric varying-coefficient partially linear models with missing responses. Firstly, the authors complement the missing response variables by regression method. Then, the empirical likelihood method is introduced to study the heteroscedasticity of the semiparametric varying-coefficient partially linear models with complete-case data. Finally, the authors obtain the finite sample property by numerical simulation.