不确定性与中国出口增长
|
摘要
本文针对近期中国出口波动幅度增加以及国际金融危机以来贸易出现"过度反应"的事实,利用中国对59个主要贸易伙伴的出口数据以及出口目标国的宏观经济变量,研究了外部不确定性与中国出口波动的数量关系和因果关系。研究发现:(1)从数量关系上来看,不确定性和外部冲击会对中国出口产生负面影响,而且这个影响的量级在短期内要远远大于传统上所考虑的供给、需求以及汇率因素所产生的影响;(2)使用灾难数据作为工具变量的两阶段最小二乘法估计表明,不确定性与中国出口增长存在着稳定的负向因果关系;(3)不确定性冲击对中国出口的影响具有异质性特征。这些发现意味着营造一个稳定有序的经济环境应该成为政策制定者的首要之选,而企业要着力对未来的需求形成科学有效的预判,以期对冲不确定性对出口增长的影响。
The recent global economic crisis saw a sharp decline in output. However,the accompanying declines in Chinese export and world trade volume were sharper and almost twice as big. The failure of traditional models to account for these developments suggests that the behavior of trade in exceptional circumstances may still be poorly understood.Traditional trade theory intrinsically fails to explain the quantitative relationship between economic uncertainty and trade,as the classical trade model is based on full information and sufficient rational assumptions. Popular analytical tools used to explain bilateral trade flows,such as the gravity equation,entirely fail to consider the effect of uncertainty.To date,the analysis of the effects of uncertainty on trade has remained largely confined to exchange rate volatility.Studies have generally found that the effect of exchange rate volatility on aggregate trade flows is "fairly small and by no means robust ". Yet,a crisis of confidence and an increase of uncertainty may easily curtail international trade,as investment and consumption decisions are put on hold( Taglioni & Zavacka,2012).Knight(1921) defines uncertainty as peoples' inability to forecast the likelihood of events happening. Uncertainty is a broad concept that can refer to fluctuations in macroeconomic variables( such as GDP, total consumption, and investment),or fluctuations in microeconomic variables( such as corporate returns,war,climate change,and economic events). Macro and micro uncertainties have typical inverse cycles. Uncertainty increases in times of economic recession and decreases as economy booms( Baker & Bloom,2013). Short-term uncertainty leads to a longer period of economic activity decline that lasts for some time. This phenomenon is referred to as the "uncertainty trap".Against both the recent increase of China's export volatility and the overreaction of trade to exogenous shocks during crises,this study investigates the statistical and causal relationship between uncertainty and China's export volatility. We construct cross-country panel data on stock market volatility as proxies for the second moments of business conditions. The study arrives at the following findings.(1) From a quantitative viewpoint,uncertainty and external shocks have a negative effect on China 's exports, and the magnitude of the effect in the short term is much larger than the traditional considerations of supply,demand,and exchange rate.(2) The two-stage least squares estimations using disaster data as instrumental variables show a stable negative causal relationship between uncertainty and China's export growth.(3) The effects of uncertainty on China's exports have heterogeneous characteristics,such as different countries,export scales,and periods. These findings mean that creating a stable and orderly economic environment should be the primary choice for policymakers,and that companies should focus on scientific and effective predictions of future demand to limit the effects of uncertainty on export growth.This study enriches and expands the current research in the following ways. First,it explains the volatility of China's total exports in relation to external uncertainty by extending specific economic variables to general confidence or psychological shocks. This differs from supply-demand and other practical economic theories used to interpret the paradigm of China's export growth. Second,the VAR model quantitatively presents the effects of the intensity and duration of uncertainty on Chinese exports,and compares the effects of uncertainty,exchange rate,and supply and demand on those exports. Third,this study helps us to better understand the potential negative effects of globalization and thus identify external risks to China's exports. Finally,this study uses natural disasters,terrorist attacks,political shocks,and other exogenous shocks as instrumental variables of uncertainty to solve potential endogenous problems.
The recent global economic crisis saw a sharp decline in output. However,the accompanying declines in Chinese export and world trade volume were sharper and almost twice as big. The failure of traditional models to account for these developments suggests that the behavior of trade in exceptional circumstances may still be poorly understood.Traditional trade theory intrinsically fails to explain the quantitative relationship between economic uncertainty and trade,as the classical trade model is based on full information and sufficient rational assumptions. Popular analytical tools used to explain bilateral trade flows,such as the gravity equation,entirely fail to consider the effect of uncertainty.To date,the analysis of the effects of uncertainty on trade has remained largely confined to exchange rate volatility.Studies have generally found that the effect of exchange rate volatility on aggregate trade flows is "fairly small and by no means robust ". Yet,a crisis of confidence and an increase of uncertainty may easily curtail international trade,as investment and consumption decisions are put on hold( Taglioni & Zavacka,2012).Knight(1921) defines uncertainty as peoples' inability to forecast the likelihood of events happening. Uncertainty is a broad concept that can refer to fluctuations in macroeconomic variables( such as GDP, total consumption, and investment),or fluctuations in microeconomic variables( such as corporate returns,war,climate change,and economic events). Macro and micro uncertainties have typical inverse cycles. Uncertainty increases in times of economic recession and decreases as economy booms( Baker & Bloom,2013). Short-term uncertainty leads to a longer period of economic activity decline that lasts for some time. This phenomenon is referred to as the "uncertainty trap".Against both the recent increase of China's export volatility and the overreaction of trade to exogenous shocks during crises,this study investigates the statistical and causal relationship between uncertainty and China's export volatility. We construct cross-country panel data on stock market volatility as proxies for the second moments of business conditions. The study arrives at the following findings.(1) From a quantitative viewpoint,uncertainty and external shocks have a negative effect on China 's exports, and the magnitude of the effect in the short term is much larger than the traditional considerations of supply,demand,and exchange rate.(2) The two-stage least squares estimations using disaster data as instrumental variables show a stable negative causal relationship between uncertainty and China's export growth.(3) The effects of uncertainty on China's exports have heterogeneous characteristics,such as different countries,export scales,and periods. These findings mean that creating a stable and orderly economic environment should be the primary choice for policymakers,and that companies should focus on scientific and effective predictions of future demand to limit the effects of uncertainty on export growth.This study enriches and expands the current research in the following ways. First,it explains the volatility of China's total exports in relation to external uncertainty by extending specific economic variables to general confidence or psychological shocks. This differs from supply-demand and other practical economic theories used to interpret the paradigm of China's export growth. Second,the VAR model quantitatively presents the effects of the intensity and duration of uncertainty on Chinese exports,and compares the effects of uncertainty,exchange rate,and supply and demand on those exports. Third,this study helps us to better understand the potential negative effects of globalization and thus identify external risks to China's exports. Finally,this study uses natural disasters,terrorist attacks,political shocks,and other exogenous shocks as instrumental variables of uncertainty to solve potential endogenous problems.
引文
陈勇兵、陈宇媚、周世民,2012:《贸易成本、企业出口动态与出口增长的二元边际》,《经济学(季刊)》第4期。
陈云松,2012:《逻辑、想象和诠释:工具变量在社会科学因果推断中的应用》,《社会学研究》第6期
李浩、胡永刚、马知遥,2007:《国际贸易与中国的实际经济周期——基于封闭与开放经济的RBC模型比较分析》,《经济研究》第5期。
毛其淋、盛斌,2013:《贸易自由化、企业异质性与出口动态——来自中国微观企业数据的证据》,《管理世界》第3期。
钱学锋、熊平,2008:《企业异质性、贸易成本与中国出口增长的二元边际》,《管理世界》第9期。
施炳展,2010:《中国出口增长的三元边际》,《经济学(季刊)》第4期。
Amiti,M.,and Weinstein,D.,2012,“Exports and Financial Shocks”,Journal of International Economics,Vol.62,277-287.
Bloom,P.N.,Scott,B.A.,and Nicholas,B.B.,2013,“Does Uncertainty Drive Business Cycles?Using Disasters as Natural Experiments”,NBER Working Paper 19475.
Barndorffnielsen,O.E.,and Shephard,N.,2006,“Econometrics of Testing for Jumps in Financial Economics Using Bipower Variation”,Journal of Financial Econometrics,Vol.4,1-30,627-675.
Bloom,N.,2009,“The Impact of Uncertainty Shocks”,Econometrica,Vol.77(3),623-685.
Bloom,N.,2014,“Fluctuations in Uncertainty”,Journal of Economic Perspectives,Vol.28(2),153-175.
Campbell,J.,Lettau,M.,Malkiel,B.,and Xu,Y.,2001,“Have Individual Stocks Become More Volatile?An Empirical Exploration of Idiosyncratic Risk”,Journal of Finance,Vol.56,1-43.
Feng,Ling,Lin,Zhiyuan,and Swenson,Deborah,2017,“Trade Policy Uncertainty and Exports:Evidence from China’s WTO Accession”,Journal of International Economics,Vol.106,20-36.
Ferrantino,M.,and Taglioni,D.,2014,“Global Value Chains in the Current Trade Slowdown”,World Bank Economic Premise,No.137,Washington,DC.
Ghironi,Fabio,and Melitz,Marc J.,2005,“International Trade and Macroeconomic Dynamics with Heterogeneous Firms”,Quarterly Journal of Economics,Vol.120,3,865-915.
Haddow,A.,Hare,C.,Hooley,J.,and Shakir,T.,2013,Macroeconomic Uncertainty:What Is It,How Can We Measure It and Why Does It Matter?Bank of England Quarterly Bulletin.
Handley,K.,and Lim2o,N.,2017,“Policy Uncertainty,Trade and Welfare:Theory and Evidence for China and the U.S.”,American Economic Review,forthcoming.
Hoover,Kevin,and Perez,Steven,1994,“Post Hoc Ergo Propter Hoc One More:An Evaluation of‘Does Monetary Policy Matter’ in the Spirit of James Tobin”,Journal of Monetary Economics,Vol.34,89-99.
Hummels,D.,and Klenow,P.,2005,“The Variety and Quality of a Nation’s Exports”,American Economic Review,Vol.95,704-23.
Jones,Ben,and Olken,Ben,2005,“Do Leaders Matter?National Leadership and Growth since World War II”,Quarterly Journal of Economics,Vol.120,835-864.
Knight,Frank H.,1921,Risk,Uncertainty,and Profit.Boston,MA:Hart,Schaffner&Marx;Houghton Miffl in Company.
Kocherlakota,N.,2009,“Some Thoughts on the State of Macro”,Available at:http://braddelong.posterous.com/narayana-kocherlakota-somethoughts-on-modern.
Novy,D.,and Taylor,A.M.,2014,“Trade and Uncertainty”,National Bureau of Economic Research.
Ravn,M.O.,and Uhlig,H.,2002,“On Adjusting the Hodrick-Prescott Filter for the Frequency of Observations.”,Review of Economics and Statistics,Vol.84,371-376.
Rodrik,D.,2011,“Unconditional Convergence”,Social Science Electronic Publishing,Vol.66(1),143-168.
Taglioni,D.,and Zavacka,V.,2012,“Innocent Bystanders:How Foreign Uncertainty Shocks Harm Exporters”,Working Paper 149,European Bank for Reconstruction and Development.
Waugh,M.E.,and Leibovici,F.,2011,“Cyclical Fluctuations in International Trade Volumes”,New York University,Working Paper.
Zavacka,Veronika,2012,“The Bullwhip Effect and the Great Trade Collapse.”,European Bank for Reconstruction and Development Working Paper,No.148.
Zussman,Assaf,Zussman,Noam,and Nielsen,Morten,2008,“Asset Market Perspectives on the Israeli-Palestinian Conflict”,Economica,Vol.75,84-115.
(1)一般用五种变量来测度不确定性:宏观变量波动、股票价格波动、汇率波动、债券收益率波动以及GDP预测偏差(GDP forecast disagreement)等(Baker&Bloom,2013)。
(2)另外,从词源上说,信贷(credit)和信心(confidence)也是一致的。Credit的拉丁词源是credo,其意思是我信任、我相信(I trust,I believe),而confidence的拉丁词源是fido,其含义也是我信任、我相信(I trust,I believe)。
(1)感谢审稿人为本文提供了这些文献。
(1)对于季度数据的HP滤波法去势,一般将λ设定为1600。对于其他观测频次,Ravn&Uhlig(2002)建议:如果是年度数据,λ应该设为6.25,如果是月度数据,λ应该设为129600。
(2)除此以外,本文还使用了两种替代方法:一是找出各个冲击事件发生期内的最大异动点作为冲击点,二是找到出现超过2个标准差的第一个月。经过检验,发现以这两种方法作为的冲击点的VAR模型与正常的VAR模型结论基本一致,从而证明了标准模型的稳健性。
(3)全球需求以剔除中国出口的全球进口总额表示,美国需求以剔除中国对美出口的美国总进口额表示。
(4)纳入这些变量主要是基于文献中对于出口决定因素的识别。其中中国总出口和对美国出口是并列关系,是本文主要分析的变量,全球需求和美国需求代表影响中国出口的需求因素,中国工业增加值代表影响中国出口的供给因素,人民币实际有效汇率的纳入主要是基于国际收支理论,而Bloom不确定性指数是基于标准普尔500指数的方差构造的,二者共同捕捉经济系统中的不确定性。
(5)该VAR模型具体包含的变量有冲击虚拟变量、中国出口总额、标准普尔500指数、人民币实际有效汇率、全球需求和中国工业增加值。
(1)图3左边的VAR模型纳入了冲击虚拟变量、中国出口总额、标准普尔500指数、人民币实际有效汇率、全球需求和中国工业增加值;右边VAR模型纳入冲击虚拟变量、中国对美国出口总额、标准普尔500指数、人民币实际有效汇率、美国需求和中国工业增加值。
(1)这是因为,好的工具变量非常难以寻觅,寻找它的逻辑和数据挖掘过程充满艰辛、难以驾驭,甚至往往需要研究者的灵感。但它在模型上的简洁性,它对社会科学想象力、逻辑力和诠释力的要求,既为定量分析提供了因果推断的重要武器,也让分析的过程充满趣味和奇思妙想(陈云松,2012)。
(2)在未做特别说明的情况下,则使用的是季度同比增长率。
(3)受篇幅所限,各个变量的详细说明并未给出,欢迎感兴趣的读者向作者来函索取。
(4)关于这些变量的详细说明可向作者索取。
(1)在经典的引力方程中,还包含了地理距离、是否接壤以及是否共同语言等变量。但是,考虑到本文回归方程的被解释变量是出口增长率,是出口绝对值经过对数差分后的数据,而以上变量基本上都是跨时不变的变量,会在差分的过程中被消掉,因此本文未将这些变量纳入回归方程。
(1)在工具变量估计中,对其外生性的检验,即所谓的过度识别检验的方法主要有两种,Sargan是在同方差的假定上的,而Hanson在异方差的基础上也可以。
(1)传统方差研究的是某个时间段内股票或组合收益率对其均值的偏离,而离差则是某一时点组合中所有股票收益率对组合平均收益率的标准偏差。一定程度上提供了投资者资产组合重置的信息,因此离差包含了市场传达的信息,能够更好预测股票回报率未来波动(Compbell et al.,2001)。
(2)囿于篇幅,本部分的估计结果和分析未能呈现,欢迎感兴趣的读者来函索取。
(3)为了保证对因果关系的有效捕捉,所有的稳健性检验仍然使用两阶段最小二乘法估计。
陈云松,2012:《逻辑、想象和诠释:工具变量在社会科学因果推断中的应用》,《社会学研究》第6期
李浩、胡永刚、马知遥,2007:《国际贸易与中国的实际经济周期——基于封闭与开放经济的RBC模型比较分析》,《经济研究》第5期。
毛其淋、盛斌,2013:《贸易自由化、企业异质性与出口动态——来自中国微观企业数据的证据》,《管理世界》第3期。
钱学锋、熊平,2008:《企业异质性、贸易成本与中国出口增长的二元边际》,《管理世界》第9期。
施炳展,2010:《中国出口增长的三元边际》,《经济学(季刊)》第4期。
Amiti,M.,and Weinstein,D.,2012,“Exports and Financial Shocks”,Journal of International Economics,Vol.62,277-287.
Bloom,P.N.,Scott,B.A.,and Nicholas,B.B.,2013,“Does Uncertainty Drive Business Cycles?Using Disasters as Natural Experiments”,NBER Working Paper 19475.
Barndorffnielsen,O.E.,and Shephard,N.,2006,“Econometrics of Testing for Jumps in Financial Economics Using Bipower Variation”,Journal of Financial Econometrics,Vol.4,1-30,627-675.
Bloom,N.,2009,“The Impact of Uncertainty Shocks”,Econometrica,Vol.77(3),623-685.
Bloom,N.,2014,“Fluctuations in Uncertainty”,Journal of Economic Perspectives,Vol.28(2),153-175.
Campbell,J.,Lettau,M.,Malkiel,B.,and Xu,Y.,2001,“Have Individual Stocks Become More Volatile?An Empirical Exploration of Idiosyncratic Risk”,Journal of Finance,Vol.56,1-43.
Feng,Ling,Lin,Zhiyuan,and Swenson,Deborah,2017,“Trade Policy Uncertainty and Exports:Evidence from China’s WTO Accession”,Journal of International Economics,Vol.106,20-36.
Ferrantino,M.,and Taglioni,D.,2014,“Global Value Chains in the Current Trade Slowdown”,World Bank Economic Premise,No.137,Washington,DC.
Ghironi,Fabio,and Melitz,Marc J.,2005,“International Trade and Macroeconomic Dynamics with Heterogeneous Firms”,Quarterly Journal of Economics,Vol.120,3,865-915.
Haddow,A.,Hare,C.,Hooley,J.,and Shakir,T.,2013,Macroeconomic Uncertainty:What Is It,How Can We Measure It and Why Does It Matter?Bank of England Quarterly Bulletin.
Handley,K.,and Lim2o,N.,2017,“Policy Uncertainty,Trade and Welfare:Theory and Evidence for China and the U.S.”,American Economic Review,forthcoming.
Hoover,Kevin,and Perez,Steven,1994,“Post Hoc Ergo Propter Hoc One More:An Evaluation of‘Does Monetary Policy Matter’ in the Spirit of James Tobin”,Journal of Monetary Economics,Vol.34,89-99.
Hummels,D.,and Klenow,P.,2005,“The Variety and Quality of a Nation’s Exports”,American Economic Review,Vol.95,704-23.
Jones,Ben,and Olken,Ben,2005,“Do Leaders Matter?National Leadership and Growth since World War II”,Quarterly Journal of Economics,Vol.120,835-864.
Knight,Frank H.,1921,Risk,Uncertainty,and Profit.Boston,MA:Hart,Schaffner&Marx;Houghton Miffl in Company.
Kocherlakota,N.,2009,“Some Thoughts on the State of Macro”,Available at:http://braddelong.posterous.com/narayana-kocherlakota-somethoughts-on-modern.
Novy,D.,and Taylor,A.M.,2014,“Trade and Uncertainty”,National Bureau of Economic Research.
Ravn,M.O.,and Uhlig,H.,2002,“On Adjusting the Hodrick-Prescott Filter for the Frequency of Observations.”,Review of Economics and Statistics,Vol.84,371-376.
Rodrik,D.,2011,“Unconditional Convergence”,Social Science Electronic Publishing,Vol.66(1),143-168.
Taglioni,D.,and Zavacka,V.,2012,“Innocent Bystanders:How Foreign Uncertainty Shocks Harm Exporters”,Working Paper 149,European Bank for Reconstruction and Development.
Waugh,M.E.,and Leibovici,F.,2011,“Cyclical Fluctuations in International Trade Volumes”,New York University,Working Paper.
Zavacka,Veronika,2012,“The Bullwhip Effect and the Great Trade Collapse.”,European Bank for Reconstruction and Development Working Paper,No.148.
Zussman,Assaf,Zussman,Noam,and Nielsen,Morten,2008,“Asset Market Perspectives on the Israeli-Palestinian Conflict”,Economica,Vol.75,84-115.
(1)一般用五种变量来测度不确定性:宏观变量波动、股票价格波动、汇率波动、债券收益率波动以及GDP预测偏差(GDP forecast disagreement)等(Baker&Bloom,2013)。
(2)另外,从词源上说,信贷(credit)和信心(confidence)也是一致的。Credit的拉丁词源是credo,其意思是我信任、我相信(I trust,I believe),而confidence的拉丁词源是fido,其含义也是我信任、我相信(I trust,I believe)。
(1)感谢审稿人为本文提供了这些文献。
(1)对于季度数据的HP滤波法去势,一般将λ设定为1600。对于其他观测频次,Ravn&Uhlig(2002)建议:如果是年度数据,λ应该设为6.25,如果是月度数据,λ应该设为129600。
(2)除此以外,本文还使用了两种替代方法:一是找出各个冲击事件发生期内的最大异动点作为冲击点,二是找到出现超过2个标准差的第一个月。经过检验,发现以这两种方法作为的冲击点的VAR模型与正常的VAR模型结论基本一致,从而证明了标准模型的稳健性。
(3)全球需求以剔除中国出口的全球进口总额表示,美国需求以剔除中国对美出口的美国总进口额表示。
(4)纳入这些变量主要是基于文献中对于出口决定因素的识别。其中中国总出口和对美国出口是并列关系,是本文主要分析的变量,全球需求和美国需求代表影响中国出口的需求因素,中国工业增加值代表影响中国出口的供给因素,人民币实际有效汇率的纳入主要是基于国际收支理论,而Bloom不确定性指数是基于标准普尔500指数的方差构造的,二者共同捕捉经济系统中的不确定性。
(5)该VAR模型具体包含的变量有冲击虚拟变量、中国出口总额、标准普尔500指数、人民币实际有效汇率、全球需求和中国工业增加值。
(1)图3左边的VAR模型纳入了冲击虚拟变量、中国出口总额、标准普尔500指数、人民币实际有效汇率、全球需求和中国工业增加值;右边VAR模型纳入冲击虚拟变量、中国对美国出口总额、标准普尔500指数、人民币实际有效汇率、美国需求和中国工业增加值。
(1)这是因为,好的工具变量非常难以寻觅,寻找它的逻辑和数据挖掘过程充满艰辛、难以驾驭,甚至往往需要研究者的灵感。但它在模型上的简洁性,它对社会科学想象力、逻辑力和诠释力的要求,既为定量分析提供了因果推断的重要武器,也让分析的过程充满趣味和奇思妙想(陈云松,2012)。
(2)在未做特别说明的情况下,则使用的是季度同比增长率。
(3)受篇幅所限,各个变量的详细说明并未给出,欢迎感兴趣的读者向作者来函索取。
(4)关于这些变量的详细说明可向作者索取。
(1)在经典的引力方程中,还包含了地理距离、是否接壤以及是否共同语言等变量。但是,考虑到本文回归方程的被解释变量是出口增长率,是出口绝对值经过对数差分后的数据,而以上变量基本上都是跨时不变的变量,会在差分的过程中被消掉,因此本文未将这些变量纳入回归方程。
(1)在工具变量估计中,对其外生性的检验,即所谓的过度识别检验的方法主要有两种,Sargan是在同方差的假定上的,而Hanson在异方差的基础上也可以。
(1)传统方差研究的是某个时间段内股票或组合收益率对其均值的偏离,而离差则是某一时点组合中所有股票收益率对组合平均收益率的标准偏差。一定程度上提供了投资者资产组合重置的信息,因此离差包含了市场传达的信息,能够更好预测股票回报率未来波动(Compbell et al.,2001)。
(2)囿于篇幅,本部分的估计结果和分析未能呈现,欢迎感兴趣的读者来函索取。
(3)为了保证对因果关系的有效捕捉,所有的稳健性检验仍然使用两阶段最小二乘法估计。