《外资利用与我国进出口贸易关系的实证分析.doc》由会员分享,可在线阅读,更多相关《外资利用与我国进出口贸易关系的实证分析.doc(16页珍藏版)》请在三一办公上搜索。
1、外资利用与我国进出口贸易关系的实证分析 摘要:利用外资与发展进出口贸易,是我国扩大对外开放过程中两个重要的方面,有效地提高了我们利用国内国外两种资源、两个市场的能力。利用外资和发展进出口贸易相互联系,相互促进。本文通过对近二十多年来我国发展实践的研究,认为二者是密不可分的。外资利用有力推动了我国进出口持续、健康的发展 ,并且外商直接投资的作用日益显著。关键词:进出口 外资 外商直接投资ABSTRACT: Utilizing foreign capital and developing export and import trade are two important aspects in ou
2、r countrys process of expanding opening to the outside world, which has effectively strengthened our ability to utilize two resources and market at home and abroad. The two aspects relate each other and promote as well. Through researching on the development practice of latest two decades of our cou
3、ntry, this article holds that they are relate closely and couldnt be separated. The utilization of foreign capital promotes the sustainable and heath development of import and export, and the effect of foreign direct investment is increasingly outstanding. KEY WORDS: export and import foreign capita
4、l foreign direct invest 目 录1 相关经济背景及问题提出31.1 我国外资利用的进程及作用31.2 问题提出32 数据的收集和整理33 模型的建立54 模型的估计和检验6 4.1 模型的估计64.2 模型的检验84.2.1 经济意义的检验84.2.2统计检验84.2.3计量经济的检验105 总结性分析15 5.1 FL对进出口的影响作用减弱15 5.2 FDI对对外贸易的影响更加显著15 5.3 FOI在外资中的比例将会上升15参考文献161 相关经济背景及研究目的1.1我国外资利用的进程及作用按照中国对外经济贸易年鉴划分,我国外资的来源主要有:对外借款、外商直接投资和
5、外商其他投资。其中,对外借款包括:政府贷款、国际金融机构贷款,国外发行债券等;外商直接投资包括:合资经营企业、合作经营企业、外资企业、外商投资股份制企业、合作开发等;外商其他投资包括:对外发行股票、国际租赁、补偿贸易、加工装配等。我国利用外资的历程分为三个阶段。第一阶段(建国至1978年),复杂的国际国内政治环境限制了我国对外资的利用,1956年以前主要是前苏联的援助,1956年中苏关系恶化,外资更是急剧减少。此阶段我国利用外资的渠道遭到封锁,外资数量极其有限。第二阶段(19791990年),中国与西方国家陆续建交,关系走上正常化。西方国家的资金开始逐步流进国门,如国外贷款,个别国外公司来华进
6、行少量投资等,虽然此阶段我国利用外资的规模是在上升,但进展却很缓慢,而且形式以接受各国政府的贷款为主,形式单一。第三阶段:1991年至今,越来越多的跨国公司来华进行规模巨大的系统投资。此阶段外商直接投资取代了对外借款,成为我国利用外资的主要形式。根据世界银行1994年度世界投资报告的统计,到1993年,我国已成为世界上继美国之后的第二大直接投资国,2002年实际利用外商直接投资逾500亿美元,首次超过美国,跃居世界第一。随着我国利用外资数量的急剧增加,我国外贸进出口以年平均12.62%的速度迅猛增长,到2003年我国进出口总额达8000亿美元,居世界第四位,其中,仅外商直接投资企业的进出口总额
7、在对外贸易总额中所占比例达56%。由此可见,利用外资对我国进出口贸易的作用巨大。1.2问题提出 已有许多学者对我国外商直接投资和进出口贸易的关系进行了实证和计量分析。陈宪,陈晨对外商直接投资和进出口额进行了相关分析,证明了外商投资增长对进出口增长有一定的拉动作用。他们研究发现外商直接投资增长与当年进出口额增长有一定的相关性,相关系数为0.423,与后1-2年进出口增长的相关性减弱,到后3-4年相关系数又开始加强。原因是外商直接投资在当年通过带动进口刺激了对外贸易增长,数年后则通过推动出口对外贸增长再次产生影响。杨巡对1980-1997年外商直接投资与我国进出口的关系进行了相关分析,发现外商直接
8、投资,特别是上一年累计外商直接投资与出口额和进口额之间都存在较高的正相关关系,相关系数分别为0.9664和0.9465,外商直接投资对我国进出口贸易有着重要的促进作用。以上这些研究主要是针对外商直接投资的研究,没有考虑外资的其他形式对进出口贸易的影响,本文基于以往研究的成果,在不考虑滞后效应的基础上全面考察外资的各种形式对进出口贸易的影响,并深入分析影响背后的原因。2 数据的收集和整理表一:我国进出口与利用外资状况 进出口贸易情况(亿美元) 实际利用外资情况(亿美元)年份进出口总额出口总额进口总额差额对外借款外商直接投资外商其他投资总计20038509.94382.34127.6254.753
9、5.046726.4841.420026207.732562951.7304527.428622.7700.120015096.526612435.5225.5468.827.9696.720004742.924922250.9241.1100407.286.4593.619993606.31949.31657292.3102.12403.1921.28526.5919983239.51837.11402.4434.7110454.6320.94585.5719973251.61827.91423.7404.2120.21452.5771.3644.0819962898.81510.51388
10、.3122.2126.69417.254.1548.0419952808.61487.81320.8167103.3375.22.9481.319942366.21210.11156.15492.67337.671.79432.1319931957917.41039.6-122.2111.89275.152.56389.619921655.3849.4805.943.579.11110.072.84192.0219911357719.1637.981.268.8843.663115.5419901154.4620.9533.587.465.334.92.7102.919891116.8525.
11、4591.4-6662.8633.923.81100.5919881027.9475.2552.7-77.564.8731.945.45102.261987826.5394.4432.1-37.758.0523.143.3384.521986738.5309.4429.1-119.750.1418.743.772.581985696273.5422.5-14925.0616.582.9844.62经过对数处理后的数据(表二):年份进出口总额出口总额进口总额对外借款外商直接投资外商其他投资总计20039.04908.38538.32556.28243.27346.330420028.73358.
12、08837.99016.26803.12246.310120018.53637.88657.79796.15023.32866.208020008.46447.82087.71914.60526.00934.45906.386219998.19047.57527.41284.62615.99943.05786.266419988.08327.51597.24594.70056.11953.04176.372619978.08697.51097.26104.78926.11494.26696.467819967.97217.32027.23584.84176.03371.41106.306319
13、957.94047.30517.18604.63765.92751.06476.176519947.76907.09857.05284.52905.82210.58226.068719937.57926.82156.94664.71755.61730.94005.965119927.41176.74456.69204.37084.70111.04385.257619917.21306.57806.45824.23243.77641.09864.749619907.05136.43126.27954.17903.55250.99334.633819897.01826.26426.38254.14
14、093.52401.33764.611119886.93536.16376.31484.17243.46391.69564.627519876.71725.97746.06874.06133.14161.20304.437019866.60465.73466.06173.91482.93071.30834.284719856.54535.61136.04623.22132.80821.09193.7982说明:利用外资数据来自中国对外经济贸易年鉴。其中,由于自2001年起,利用外资数据中的对外借款部分不再统计,造成了部分数据缺失,同时对外借款属于主观行为,基本无规律可循,不能通过回归等方法进行
15、估计,所以,本文只采取19852000年的数据。值得说明的是:1997年起对外发行股票从对外借款中剔除,列入外商其他投资额中。本文假设数据统计的变化不对分析结果产生影响。对数数据的取得是利用EXCEL软件计算得到。3 模型的建立下面分别建立利用外资的各个部分对我国出口、进口和进出口总额的影响的计量经济模型:lnEX=0+1lnFDI+2lnFL+3lnFOI+t,其中:lnEX取对数后的出口lnFDI取对数后的外商直接投资lnFL取对数后的对外借款lnFOI取对数后的外商其他投资0、1、2、3均为待定系数为随机误差项 lnIM=0+1lnFDI+2lnFL+3lnFOI+et,其中:lnIM取
16、对数后的进口lnFDI取对数后的外商直接投资lnFL取对数后的对外借款lnFOI取对数后的外商其他投资0、1、2、3均为待定系数e为随机误差项lnXM=0+1lnFDI+2lnFL+3lnFOI+,其中:lnXM取对数后的进出口总额lnFDI取对数后的外商直接投资lnFL取对数后的对外借款lnFOI取对数后的外商其他投资0、1、2、3均为待定系数,为随机误差项4 模型的估计和检验4.1模型的估计用Eviews对19852000年的数据进行多元回归得到:Dependent Variable: LNEXMethod: Least SquaresDate: 05/26/05 Time: 12:05S
17、ample: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C3.6087790.7357214.9050910.0004LNFDI0.3679070.0770384.7756620.0005LNFL0.2793930.2355081.1863410.2584LNFOI0.1208360.0423572.8527980.0145R-squared0.951497 Mean dependent var6.779568Adjusted R-squared0.939371 S.D. de
18、pendent var0.701176S.E. of regression0.172650 Akaike info criterion-0.462786Sum squared resid0.357695 Schwarz criterion-0.269639Log likelihood7.702289 F-statistic78.46923Durbin-Watson stat1.242766 Prob(F-statistic)0.000000 Dependent Variable: LNIMMethod: Least SquaresDate: 05/26/05 Time: 12:07Sample
19、: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C5.2219670.49800110.485870.0000LNFDI0.3785650.0521467.2597100.0000LNFL-0.0912390.159412-0.5723440.5777LNFOI0.0901340.0286713.1437280.0085R-squared0.963592 Mean dependent var6.772717Adjusted R-squared0.954490 S.D. depen
20、dent var0.547809S.E. of regression0.116864 Akaike info criterion-1.243286Sum squared resid0.163888 Schwarz criterion-1.050139Log likelihood13.94629 F-statistic105.8659Durbin-Watson stat1.217006 Prob(F-statistic)0.000000 Dependent Variable: LNXMMethod: Least SquaresDate: 05/26/05 Time: 12:08Sample: 1
21、985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C5.1896440.5517379.4060170.0000LNFDI0.3771030.0577736.5273500.0000LNFL0.0718000.1766130.4065390.6915LNFOI0.1067860.0317653.3618020.0057R-squared0.964980 Mean dependent var7.473900Adjusted R-squared0.956225 S.D. dependent
22、var0.618827S.E. of regression0.129475 Akaike info criterion-1.038347Sum squared resid0.201164 Schwarz criterion-0.845200Log likelihood12.30677 F-statistic110.2193Durbin-Watson stat1.110469 Prob(F-statistic)0.000000写出回归分析报告:lnEX=3.608779+0.367907lnFDI+0.279393lnFL+0.120836lnFOI 0.735721 0.077038 0.23
23、5508 0.042357 4.905091 4.775662 1.186341 2.852798R2=0.951497 修正后R2=0.939371 F=78.46923 DW=1.242766lnIM=5.221967+0.378565lnFDI-0.091239lnFL+0.090134lnFOI0.498001 0.052146 0.159412 0.028671 10.48587 7.259710 -0.572344 3.143728R2=0.963592 修正后R2=0.954490 F=105.8659 DW=1.217006lnXM=5.189644+0.377103lnFDI
24、+0.071800lnFL+0.106786lnFOI 0.551737 0.057773 0.176613 0.031765 9.406017 6.527350 0.406539 3.143728R2=0.964980 修正后R2=0.956225 F=110.2193 DW=1.1104694.2模型的检验:4.2.1经济意义的检验第一,FDI对EX、IM、XM的影响。回归结果显示:FDI每增加1%,EX、IM、XM分别增加0.367907%、0.378565%、0.377103%。第二,FL对EX、IM、XM的影响。回归结果显示:FL每增加1%,EX、IM、XM分别增加0.279393%
25、、-0.091239%、0.071800%。第三,FOI对EX、IM、XM的影响。回归结果显示:FOI每增加1%,EX、IM、XM分别增加0.120836%、0.090134%、0.106786%。FDI、FL、FOI对EX、XM的影响都表现为正的拉动作用,FDI、FOI对IM的影响也表现为正的拉动作用,仅FL对IM的影响表现为负的抑制作用。并且在外资的三种表现因素中,FDI对EX、IM、XM的影响最大,其次是FOI,FL的影响最小。4.2.2统计检验三个模型R2、修正后R2值都较大(都在90%以上),说明模型的样本回归线对样本的拟合优度好。以模型一为例,R2=0.951497,修正后R2=0
26、.939371说明总离差平方和的95.1497%(修正后93.9371%)被样本回归线解释,其余部分不能被解释变量解释。 查表得:F0.05(3,12) =3.49,显然三个模型的F统计量值远远大于临界值,全部通过F检验。 查表得:t0.05(12)=2.179,所以,LnFDI(4.775662、7.259710、6.527350)、LnFOI(2.852798、3.143728、3.143728)的系数通过t检验,而LnFL的系数(1.186341、-0.572344、0.406539)没有通过t检验。由于LnFL的系数不显著,我们剔除变量LnFL,重新进行回归得到:Dependent V
27、ariable: LNEXMethod: Least SquaresDate: 05/26/05 Time: 13:35Sample: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C4.4598260.16583226.893600.0000LNFDI0.4472200.03887511.504000.0000LNFOI0.1165230.0428572.7188880.0176R-squared0.945808 Mean dependent var6.779568Adjuste
28、d R-squared0.937471 S.D. dependent var0.701176S.E. of regression0.175334 Akaike info criterion-0.476886Sum squared resid0.399647 Schwarz criterion-0.332025Log likelihood6.815085 F-statistic113.4449Durbin-Watson stat0.950502 Prob(F-statistic)0.000000Dependent Variable: LNIMMethod: Least SquaresDate:
29、05/26/05 Time: 13:35Sample: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C4.9440480.10763545.933610.0000LNFDI0.3526650.02523213.976800.0000LNFOI0.0915420.0278173.2909200.0058R-squared0.962598 Mean dependent var6.772717Adjusted R-squared0.956844 S.D. dependent var0.
30、547809S.E. of regression0.113802 Akaike info criterion-1.341354Sum squared resid0.168361 Schwarz criterion-1.196493Log likelihood13.73083 F-statistic167.2880Durbin-Watson stat1.222392 Prob(F-statistic)0.000000Dependent Variable: LNXMMethod: Least SquaresDate: 05/26/05 Time: 13:36Sample: 1985 2000Inc
31、luded observations: 16VariableCoefficientStd. Errort-StatisticProb. C5.4083520.11846145.655050.0000LNFDI0.3974860.02777014.313390.0000LNFOI0.1056780.0306153.4518920.0043R-squared0.964497 Mean dependent var7.473900Adjusted R-squared0.959035 S.D. dependent var0.618827S.E. of regression0.125249 Akaike
32、info criterion-1.149668Sum squared resid0.203935 Schwarz criterion-1.004808Log likelihood12.19734 F-statistic176.5848Durbin-Watson stat1.041695 Prob(F-statistic)0.000000通过统计检验的回归分析报告:lnEX=4.459826+0.447220lnFDI+0.116523lnFOI 0.165832 0.038875 0.042857 26.89361 11.50400 2.718888R2=0.945808 修正后R2=0.93
33、7471 F=113.4449 DW=0.950502lnIM=4.944048+0.352665lnFDI+0.091542lnFOI0.107635 0.025232 0.027817 45.93361 13.97680 3.290920 R2=0.962598 修正后R2=0.956844 F=167.2880 DW=1.222392lnXM=5.408352+0.397486lnFDI+0.105678lnFOI 0.118461 0.027770 0.030615 45.65505 14.31339 3.451892R2=0.964497 修正后R2=0.959035 F=176.5
34、848 DW=1.0416954.2.3计量经济的检验第一、多重共线性的检验在我们的模型中只剩两个解释变量LnFDI、LnFOI,只要检验它们之间的共线性即可。首先,采用简单相关系数矩阵法,可得到两个解释变量之间的相关系数r =0.497073。这说明两变量之间存在一定的相关性,但相关性不强。为进一步验证两者之间的共线性,我们采用辅助回归进行进一步的检验。其次,才用辅助回归法将作为LnFDI被解释变量,LnFOI作为解释变量,回归结果如下:Dependent Variable: LNFDIMethod: Least SquaresDate: 05/26/05 Time: 13:53Sample
35、: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C3.7420100.5473406.8367140.0000LNFOI0.5479850.2556572.1434360.0501R-squared0.247082 Mean dependent var4.721374Adjusted R-squared0.193302 S.D. dependent var1.342070S.E. of regression1.205399 Akaike info criterion3.32796
36、7Sum squared resid20.34181 Schwarz criterion3.424540Log likelihood-24.62373 F-statistic4.594319Durbin-Watson stat0.271176 Prob(F-statistic)0.050126R2=0.247082,修正后R2=0.193302,t0.05(14)=2.1452.143436,F0.05(1,14) =4.604.594319。说明LnFDI、LnFOI之间不存在共线性,我们仍然保持模型不变。第二、异方差性的检验我们采用White(no cross)检验,结果如下:White
37、Heteroskedasticity Test:F-statistic0.807205 Probability0.545808Obs*R-squared3.630739 Probability0.458280Test Equation:Dependent Variable: RESID2Method: Least SquaresDate: 05/26/05 Time: 14:15Sample: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C0.0290590.2196310.13
38、23060.8971LNFDI0.0285700.0987380.2893540.7777LNFDI2-0.0041200.010876-0.3788070.7120LNFOI-0.0474520.040369-1.1754770.2646LNFOI20.0097370.0078351.2427670.2398R-squared0.226921 Mean dependent var0.024978Adjusted R-squared-0.054198 S.D. dependent var0.030297S.E. of regression0.031107 Akaike info criteri
39、on-3.852454Sum squared resid0.010644 Schwarz criterion-3.611020Log likelihood35.81963 F-statistic0.807205Durbin-Watson stat2.283545 Prob(F-statistic)0.545808White Heteroskedasticity Test:F-statistic54.28333 Probability0.000000Obs*R-squared15.22852 Probability0.004250Test Equation:Dependent Variable:
40、 RESID2Method: Least SquaresDate: 05/26/05 Time: 14:12Sample: 1985 2000Included observations: 16VariableCoefficientStd. Errort-StatisticProb. C0.0156720.0341210.4593070.6550LNFDI0.0025740.0153390.1678300.8698LNFDI2-0.0004340.001690-0.2570300.8019LNFOI-0.0226570.006271-3.6127890.0041LNFOI20.0073310.0012176.0230070.0001R-squared0.951783 Mean dependent var0.010523Adjusted R-squared0.934249 S.D. dependent var0.018846S.E. of regression0.004833