中国混频金融状况指数的经济增长预测效果与检验
|
摘要
基于TVP-FAVAR模型及修订式测算中国FCI,并刻画SF-FCI和MF-FCI的经济增长预测效果与冲击效应,进一步运用混频Granger方法检验FCI与经济增长之间的因果关系,结果表明:合成FCI的各金融变量缺口存在较大差异,简单同频转换存在一定误差;SF-FCI领先经济增长大约1～2季度,MF-FCI领先经济增长大约2～4个季度,MF-FCI的经济增长效应与拟合效果更好,"典型时期"预测经济增长的先导性更强;SF-FCI和MF-FCI是经济增长的混频Granger原因,经济增长不是SF-FCI的混频Granger原因;经济增长是MF-FCI的混频Granger原因,混频Granger检验下MF-FCI和经济增长的因果关系更密切。
This article calculates China's financial conditions index based on the TVP-FAVAR model and revised form,which emphasizes the economic growth forecasting and shocking effects of the same frequency financial conditions index and mixed financial conditions index,and further uses the mixed frequency Granger method to test the relationship between the financial conditions index and economic growth.The results show that:Gap of the financial variables which synthesizes FCI has big difference,the simple same frequency conversion has a certain error;SF-FCI keeps ahead economic growth about 1～2quarters,MF-FCI keeps ahead economic growth about 2～4quarters,economic growth effect prediction ability of MF-FCI are better,MF-FCI predicts stronger economic growth in the typical period;SF-FCI and MF-FCI are the mixed frequency Granger causes of economic growth,economic growth is not the mixed frequency Granger cause of SF-FCI,economic growth is the mixed frequency Granger cause of MF-FCI,and the causality relationship between MF-FCI and economic growth is closer under mixed frequency Granger causality test.
This article calculates China's financial conditions index based on the TVP-FAVAR model and revised form,which emphasizes the economic growth forecasting and shocking effects of the same frequency financial conditions index and mixed financial conditions index,and further uses the mixed frequency Granger method to test the relationship between the financial conditions index and economic growth.The results show that:Gap of the financial variables which synthesizes FCI has big difference,the simple same frequency conversion has a certain error;SF-FCI keeps ahead economic growth about 1～2quarters,MF-FCI keeps ahead economic growth about 2～4quarters,economic growth effect prediction ability of MF-FCI are better,MF-FCI predicts stronger economic growth in the typical period;SF-FCI and MF-FCI are the mixed frequency Granger causes of economic growth,economic growth is not the mixed frequency Granger cause of SF-FCI,economic growth is the mixed frequency Granger cause of MF-FCI,and the causality relationship between MF-FCI and economic growth is closer under mixed frequency Granger causality test.
引文
[1]蔡晓春,郭玉鑫.政府干预、金融发展对贫困减缓的影响分析——基于非线性交互效应动态面板模型[J].统计与信息论坛,2018(9).
[2] Freedman C.The Use of Indicators and the Monetary Conditions Index in Canada[R].Frameworks of Monetary Stability:Policy Issues and Country Experiments,IMF Working Paper,1994.
[3] Ball L.Efficient Rules for Monetary Policy[J].International Finance,1999,2(1).
[4] Goodhart C,Hofmann B.Asset Prices and the Conduct of Monetary Policy[J].Royal Economic Society,2002(8).
[5] Van den End J W.Indicator and Boundaries of Financial Stability[R].Amsterdam:De Nederlandsche Bank Working Paper,De Nederlandsche Bank,2006.
[6] Matheson T D.Financial Conditions Indexes for the United States and Euro Area[J].Economics Letters,2012,115(3).
[7] Koop G,Korobilis D.A New Index of Financial Conditions[J].European Economic Review,2014,71.
[8]周德才,朱志亮,刘琪,王婧薇.中国FCI构建及其对通胀的非对称性效应——基于BDFA-VAR模型和MS-VAR模型的实证分析[J].统计与信息论坛,2017(8).
[9]邓创,滕立威,徐曼.中国金融状况的波动特征及其宏观经济效应分析[J].国际金融研究,2016(3).
[10]栾惠德,侯晓霞.中国实时金融状况指数的构建[J].数量经济技术经济研究,2015(4).
[11]尚玉皇,郑挺国.中国金融形势指数混频测度及其预警行为研究[J].金融研究,2018(3).
[12]刘汉,王永莲,陈德凯.混频Granger因果关系检验的功效和稳健性分析[J].数量经济技术经济研究,2017(10).
[13]张辉,黄泽华.我国货币政策的汇率传导机制研究[J].经济学动态,2011(8).
[14]贾庆英,孔艳芳.资产价格、经济杠杆与价格传递——基于国际PVAR模型的实证研究[J].国际金融研究,2016(1).
[15]姚余栋,李法瑾.中国货币政策传导信贷渠道的经验研究:总量融资结构的新证据[J].世界经济,2013(3).
[16]刘金全,徐宁,刘达禹.资产价格错位与货币政策规则——基于修正Q理论的重新审视[J].国际金融研究,2017(5).
[17]郭娜,翟光宇.中国利率政策与房地产价格的互动关系研究[J].经济评论,2011(3).
(1)利率和汇率实际可以看成资产价格,通过资产价格传导渠道进行分析,由于利率、汇率的传导渠道研究较为成熟;本文单独推演,文中资产价格传导主要指股票价格传导、房地产价格传导。
(2)考虑到金融变量在数据类型和计量单位存在差异,为了避免不同缺口值在量纲上的差异,且使回归系数及变量权重具有可比性,需对缺口序列进行标准化处理,使其在技术上具有可比性和可加性。
(1)类季度数据的选择有81种方案,本文选择其中一种序列构建方式予以说明,其他方案并未一一列举。此外,对于各金融变量的动态权重及其政策导向原因,并非本文研究重点,故并未做具体分析。
(1)带有随机波动率的TVP-FAVAR模型记为SV-TVP-FAVAR模型,此处不做详细介绍。
[2] Freedman C.The Use of Indicators and the Monetary Conditions Index in Canada[R].Frameworks of Monetary Stability:Policy Issues and Country Experiments,IMF Working Paper,1994.
[3] Ball L.Efficient Rules for Monetary Policy[J].International Finance,1999,2(1).
[4] Goodhart C,Hofmann B.Asset Prices and the Conduct of Monetary Policy[J].Royal Economic Society,2002(8).
[5] Van den End J W.Indicator and Boundaries of Financial Stability[R].Amsterdam:De Nederlandsche Bank Working Paper,De Nederlandsche Bank,2006.
[6] Matheson T D.Financial Conditions Indexes for the United States and Euro Area[J].Economics Letters,2012,115(3).
[7] Koop G,Korobilis D.A New Index of Financial Conditions[J].European Economic Review,2014,71.
[8]周德才,朱志亮,刘琪,王婧薇.中国FCI构建及其对通胀的非对称性效应——基于BDFA-VAR模型和MS-VAR模型的实证分析[J].统计与信息论坛,2017(8).
[9]邓创,滕立威,徐曼.中国金融状况的波动特征及其宏观经济效应分析[J].国际金融研究,2016(3).
[10]栾惠德,侯晓霞.中国实时金融状况指数的构建[J].数量经济技术经济研究,2015(4).
[11]尚玉皇,郑挺国.中国金融形势指数混频测度及其预警行为研究[J].金融研究,2018(3).
[12]刘汉,王永莲,陈德凯.混频Granger因果关系检验的功效和稳健性分析[J].数量经济技术经济研究,2017(10).
[13]张辉,黄泽华.我国货币政策的汇率传导机制研究[J].经济学动态,2011(8).
[14]贾庆英,孔艳芳.资产价格、经济杠杆与价格传递——基于国际PVAR模型的实证研究[J].国际金融研究,2016(1).
[15]姚余栋,李法瑾.中国货币政策传导信贷渠道的经验研究:总量融资结构的新证据[J].世界经济,2013(3).
[16]刘金全,徐宁,刘达禹.资产价格错位与货币政策规则——基于修正Q理论的重新审视[J].国际金融研究,2017(5).
[17]郭娜,翟光宇.中国利率政策与房地产价格的互动关系研究[J].经济评论,2011(3).
(1)利率和汇率实际可以看成资产价格,通过资产价格传导渠道进行分析,由于利率、汇率的传导渠道研究较为成熟;本文单独推演,文中资产价格传导主要指股票价格传导、房地产价格传导。
(2)考虑到金融变量在数据类型和计量单位存在差异,为了避免不同缺口值在量纲上的差异,且使回归系数及变量权重具有可比性,需对缺口序列进行标准化处理,使其在技术上具有可比性和可加性。
(1)类季度数据的选择有81种方案,本文选择其中一种序列构建方式予以说明,其他方案并未一一列举。此外,对于各金融变量的动态权重及其政策导向原因,并非本文研究重点,故并未做具体分析。
(1)带有随机波动率的TVP-FAVAR模型记为SV-TVP-FAVAR模型,此处不做详细介绍。