金融周期对货币政策有效性的影响研究:警惕金融繁荣下的货币政策效力扭曲
|
摘要
综合金融市场的多维信息,利用主成分分析法和BP滤波法对我国金融周期进行客观、系统地测度。在此基础上,探讨金融周期对中国货币政策有效性的影响。研究结果表明:我国金融周期具有中周期性质,周期长度约为9年,振幅有缩小趋势,由0.034缩小至0.016;金融中周期对货币政策有效性影响显著,金融繁荣阶段,货币政策产出效应下降约50%,价格效应则至少增强1倍;金融繁荣下产出效应的下降将加大货币政策调控成本,增加政策无效的风险。由此,本文认为将金融周期的中周期特征纳入货币政策视野存在必要性,货币当局应警惕金融繁荣下的货币政策效力扭曲,审慎操作,且不宜承担过多的刺激产出的任务。
Based on the multi-dimensional information of the financial market,the principal component analysis method and BP filtering method were used to measure the financial cycle of China.On this basis,the paper explored the impact of the financial cycle on the effectiveness of China's monetary policy.The results show that China's financial cycle has a medium-cycle nature,with a period length of about 9years,and the amplitude has a narrowing trend,which is reduced from 0.034 to 0.016.The financial cycle has a significant impact on the effectiveness of monetary policy:the output effect of monetary policy declines almost 50% whereas the price effect is at least doubled during the financial boom.Financial prosperity will increase the cost and pressure of monetary policy regulation output and increase the risk of policy ineffectiveness.Therefore,this paper believes that it is necessary to incorporate the mid-cycle characteristics of the financial cycle into the monetary policy vision.The monetary authorities should be wary of the monetary policy effectiveness distortion under the financial prosperity,prudent operation,and should not undertake too many tasks to stimulate output.
Based on the multi-dimensional information of the financial market,the principal component analysis method and BP filtering method were used to measure the financial cycle of China.On this basis,the paper explored the impact of the financial cycle on the effectiveness of China's monetary policy.The results show that China's financial cycle has a medium-cycle nature,with a period length of about 9years,and the amplitude has a narrowing trend,which is reduced from 0.034 to 0.016.The financial cycle has a significant impact on the effectiveness of monetary policy:the output effect of monetary policy declines almost 50% whereas the price effect is at least doubled during the financial boom.Financial prosperity will increase the cost and pressure of monetary policy regulation output and increase the risk of policy ineffectiveness.Therefore,this paper believes that it is necessary to incorporate the mid-cycle characteristics of the financial cycle into the monetary policy vision.The monetary authorities should be wary of the monetary policy effectiveness distortion under the financial prosperity,prudent operation,and should not undertake too many tasks to stimulate output.
引文
[1]Fisher I.The Debt-Deflation Theory of Great Depressions[J].Econometrica,1933,1(4):337-357.
[2]Minsky H P.The Financial Instability Hypothesis:An Interpretation of Keynes and an Alternative to“Standard”Theory[J].Challenge,1977,20(1):20-27.
[3]Bernanke B,Gertler M,Gilchrist S.Handbook of Macroeconomics[M].Elsevier,1999:1341-1393.
[4]Kiyotaki N,Moore J.Credit Cycles[J].Journal of Political Economy,1997,105(2):211-248.
[5]Claessens S,Kose M A,Terrones M E.How do Business and Financial Cycles Interact?[J].Journal of International Economics,2012,87(1):178-190.
[6]Drehmann M,Borio C E V,Tsatsaronis K.Characterising the Financial Cycle:Don’t Lose Sight of the Medium Term!Bis Working Papers[R].The Bank for International Settlements,2012,68(3).
[7]Borio C.The Financial Cycle and Macroeconomics:What have We Learnt?[J].Journal of Banking and Finance,2014,45:182-198.
[8]范小云,袁梦怡,肖立晟.理解中国的金融周期:理论、测算与分析[J].国际金融研究,2017,357(1).
[9]韩艾,郑桂环,汪寿阳.广义动态因子模型在景气指数构建中的应用---中国金融周期景气分析[J].系统工程理论与实践,2010,30(5).
[10]徐国祥,郑雯.中国金融状况指数的构建及预测能力研究[J].统计研究,2013,30(8).
[11]王博,李昊然.中国金融周期测度及国际比较研究[J].经济学动态,2018(6).
[12]Borio C,Zhu H.Capital Regulation,Risk-Taking and Monetary Policy:A Missing Link in the Transmission Mechanism?[J].Journal of Financial Stability,2008,8(4):236-251.
[13]邓创,徐曼.中国的金融周期波动及其宏观经济效应的时变特征研究[J].数量经济技术经济研究,2014,31(9).
[14]Borio C,Disyatat P,Juselius M.Rethinking Potential Output:Embedding Information about the Financial Cycle.Bis Working Papers[R].The Bank for International Settlements,2013,69(3).
[15]张成思,张步昙.中国实业投资率下降之谜:经济金融化视角[J].经济研究,2016,51(12).
[16]刘玚.基于金融稳定因素的中国货币政策传导效应研究[J].金融经济学研究,2016,31(1).
[17]张晓晶,王宇.金融周期与创新宏观调控新维度[J].经济学动态,2016(7).
[18]马勇,张靖岚,陈雨露.金融周期与货币政策[J].金融研究,2017(3).
[19]裴平,熊鹏,朱永利.经济开放度对中国货币政策有效性的影响:基于1985-2004年交叉数据的分析[J].世界经济,2006(5).
[20]王玉.资本账户开放程度、融资约束对货币政策有效性的影响[J].金融理论与实践,2018(5).
[21]Berument H,Konac N,Senay O.Openness and the Effectiveness of Monetary Policy:A Cross-Country Analysis[J].International Economic Journal,2007,21(4):577-591.
(1)从主成分分析的结果来看,所选指标的KMO检验与Bartlett球度检验结果均符合要求。因此,选择累计贡献率达78.389%的前四个主成分,以各自方差贡献率为权数合成我国的金融运行指数(FI),公式如下:FI=0.351 7×F1+0.217 35×F2+0.127 37×F3+0.087 4×F4
(1)由于lnY,lnP,lnM均为一阶单整序列且三者之间存在协整关系。故可以利用各变量的一阶差分建立ARMAX模型对未预期到的货币冲击进行估计。解释变量与自回归模型的滞后阶数由t统计量和AIC、BIC信息准则综合判定。根据货币供给方程的性质,最终模型ARMAX(5.,1,0)的残差项表示满足经济增长、物价变化之外的货币供给冲击,即本文的未预期到的货币冲击μt。
(1)模型1、3、5、7为包含全部变量的未受约束模型的实证结果;模型2、4、6、8为去除不显著变量后,受约束模型的实证结果。显然,二者结论一致,且受约束模型的检验结果更好。因此,本文的主要结论主要基于后者。
(1)如前文的测算结果,2003年至今我国经历了约1.5个金融周期。为了使每个子样本包含足够的金融周期变量,同时克服了不同样本数量对结果的影响,故将子样本划分为2004-2013年和2007-2017年两个子样本。事实上,文本也对2004-2008年子样本进行了回归检测,结果依然支持本文的结论。
[2]Minsky H P.The Financial Instability Hypothesis:An Interpretation of Keynes and an Alternative to“Standard”Theory[J].Challenge,1977,20(1):20-27.
[3]Bernanke B,Gertler M,Gilchrist S.Handbook of Macroeconomics[M].Elsevier,1999:1341-1393.
[4]Kiyotaki N,Moore J.Credit Cycles[J].Journal of Political Economy,1997,105(2):211-248.
[5]Claessens S,Kose M A,Terrones M E.How do Business and Financial Cycles Interact?[J].Journal of International Economics,2012,87(1):178-190.
[6]Drehmann M,Borio C E V,Tsatsaronis K.Characterising the Financial Cycle:Don’t Lose Sight of the Medium Term!Bis Working Papers[R].The Bank for International Settlements,2012,68(3).
[7]Borio C.The Financial Cycle and Macroeconomics:What have We Learnt?[J].Journal of Banking and Finance,2014,45:182-198.
[8]范小云,袁梦怡,肖立晟.理解中国的金融周期:理论、测算与分析[J].国际金融研究,2017,357(1).
[9]韩艾,郑桂环,汪寿阳.广义动态因子模型在景气指数构建中的应用---中国金融周期景气分析[J].系统工程理论与实践,2010,30(5).
[10]徐国祥,郑雯.中国金融状况指数的构建及预测能力研究[J].统计研究,2013,30(8).
[11]王博,李昊然.中国金融周期测度及国际比较研究[J].经济学动态,2018(6).
[12]Borio C,Zhu H.Capital Regulation,Risk-Taking and Monetary Policy:A Missing Link in the Transmission Mechanism?[J].Journal of Financial Stability,2008,8(4):236-251.
[13]邓创,徐曼.中国的金融周期波动及其宏观经济效应的时变特征研究[J].数量经济技术经济研究,2014,31(9).
[14]Borio C,Disyatat P,Juselius M.Rethinking Potential Output:Embedding Information about the Financial Cycle.Bis Working Papers[R].The Bank for International Settlements,2013,69(3).
[15]张成思,张步昙.中国实业投资率下降之谜:经济金融化视角[J].经济研究,2016,51(12).
[16]刘玚.基于金融稳定因素的中国货币政策传导效应研究[J].金融经济学研究,2016,31(1).
[17]张晓晶,王宇.金融周期与创新宏观调控新维度[J].经济学动态,2016(7).
[18]马勇,张靖岚,陈雨露.金融周期与货币政策[J].金融研究,2017(3).
[19]裴平,熊鹏,朱永利.经济开放度对中国货币政策有效性的影响:基于1985-2004年交叉数据的分析[J].世界经济,2006(5).
[20]王玉.资本账户开放程度、融资约束对货币政策有效性的影响[J].金融理论与实践,2018(5).
[21]Berument H,Konac N,Senay O.Openness and the Effectiveness of Monetary Policy:A Cross-Country Analysis[J].International Economic Journal,2007,21(4):577-591.
(1)从主成分分析的结果来看,所选指标的KMO检验与Bartlett球度检验结果均符合要求。因此,选择累计贡献率达78.389%的前四个主成分,以各自方差贡献率为权数合成我国的金融运行指数(FI),公式如下:FI=0.351 7×F1+0.217 35×F2+0.127 37×F3+0.087 4×F4
(1)由于lnY,lnP,lnM均为一阶单整序列且三者之间存在协整关系。故可以利用各变量的一阶差分建立ARMAX模型对未预期到的货币冲击进行估计。解释变量与自回归模型的滞后阶数由t统计量和AIC、BIC信息准则综合判定。根据货币供给方程的性质,最终模型ARMAX(5.,1,0)的残差项表示满足经济增长、物价变化之外的货币供给冲击,即本文的未预期到的货币冲击μt。
(1)模型1、3、5、7为包含全部变量的未受约束模型的实证结果;模型2、4、6、8为去除不显著变量后,受约束模型的实证结果。显然,二者结论一致,且受约束模型的检验结果更好。因此,本文的主要结论主要基于后者。
(1)如前文的测算结果,2003年至今我国经历了约1.5个金融周期。为了使每个子样本包含足够的金融周期变量,同时克服了不同样本数量对结果的影响,故将子样本划分为2004-2013年和2007-2017年两个子样本。事实上,文本也对2004-2008年子样本进行了回归检测,结果依然支持本文的结论。