题目：汇率和外汇储备的变化对进出口总额的影响

计量经济学期末考试论文题目：汇率和外汇储备的变化对进出口总额的影响姓 名:银聪学 号：130713982班 级：会计项目五班指导教师：米国芳时 间：二O六年一月六日汇率和外汇储备的变化对进出口总额的影响。一、研究背景汇率和外汇储备是一国国际储备的重要组成部分，它对进出口总额有重要 的作用。汇率变动、外汇储备不论是储备货币本身价值的变化，还是本国 货币汇率的变化都会对一国的出口总额产生影响汇率变动、增加或减 少外汇储备所代表的实际价值，都会对进出口总额有影响。二实验过程（一）模型设定（Y、XI、X2）为了研究我国汇率和外汇储备的变化对进出口总额的影响。，贝U理论模型设定为：1）其中，y表示进出口总额，XI表示汇率，X2表示外汇储备。由19852014年我国进出口总憑人民币亿元）Y，汇率（人民币对美元）X1,外汇储备（亿美元）X2。数据区间：19852014年进出口总额的因素年份进出口总额（人民币亿元）Y汇率（人民币对美元）X1外汇储备（亿美元）X219852066.70293.6626.4419862580.40345.2820.7219873084.20372.2129.2319883821.80372.2133.7219894155.90376.5155.5019905560.10478.32110.9319917225.80532.33217.1219929119.60551.46194.43199311271.00576.20211.99199420381.90861.87516.20199523499.90835.10735.97199624133.80831.421050.29199726967.20828.981398.9199826849.70827.911449.59199929896.30827.831546.75200039273.20827.841655.74200142183.60827.702121.65200251378.20827.702864.07200370483.50827.704032.51200495538.90827.686099.322005116921.80819.178188.722006140974.06797.1810663.42007166863.70760.4015282.492008179921.47694.5119460.32009150648.06683.1023991.522010201727.14676.9528473.382011236401.95645.8831811.482012244160.30631.2533115.892013258168.90619.3238213.152014264334.50614.2838430.18表一（数据来源:统计年鉴）（二）参数估计（OLS估计，包含参数置信区间）1、数据的计算机输入及运行过程、模型的结果（1）在EVIEWS中新建工作簿，定义变量（x1）、（x2）及（y）, 并输入相关数据，通过EVIEWS软件计算，得出估计模型的参数结果 如下：File Edit Object View Proc Quick Options Add-ins Window Helpdata yx1 x2ViewRan；SamZl Workfile: UNTITLED Group: UNTITLED Workfile: UNTITLED:Untitled- n x| View | Proc | Object | | Print | Name | Freeze | Default v | Sort | Transpose | Edit*/-1 SmplobsYX1X2obsYX1X219852066.700293.660026.4400019862580.400345.280020.7200019873084.200372.210029.2300019883821.800372.210033.7200019894155.900376.510055.5000019905560.100478.3200110.930019917225.800532.3300217.120019929119.600551.4600194.4300199311271.00576.2000211.9900199420381.90861.8700516.2000199523499.90835.1000735.9700199624133.80831.42001050.290199726967.20828.98001398.900199826849.70827.91001449.590199929896.30827.83001546.750200039273.20827.84001655.740200142183.60827.70002121.650200251378.20827.70002864.070200370483.50827.70004032.510200495538.90827.68006099.3202005116921.8819.17008188.7202006rcrFile Edit Object View Proc Quick Options Add-ins Window Help data yx1 x2Is y cx1 x20 Workfile: UNTITLED Group: UNTITLED Workfile: UNTITLED:Untitled- n x(S0M0MLView | Pro2066.7obs1985 198619871986 1989J19901991199219931994 19951 1996 19971199819992000 200120022003200420052006(=Equation: UNTITLED Workfile: UNTITLED:Untitled - n x | View I Proc Object | | Print NameFreeze | Estimate | Forecast Stats | Res ids Dependent Variable: YMethod: Least SquaresDate: 01/06/16 Time: 14:33Sample: 1985 2014In eluded observatio ns: 30VariableCoefficientStd. Errort-StatisticProb.C-24628.7112610.57-1.9530210.0613X169.0725618.153883.8048370.0007X26.6821660.2522仃26.493680.0000R-squared0.964257Mean dependentvar81986.45Adjusted R-squared0.961609S.D.dependent var89406.83S.E. of regression17517.93Akaike info criterion22.47448Sum squared resid8.29E+09Schwarz criterio n22.61460Log likelihood-334.1172Hannan-Quinn criter.22.51930F-statistic364.1973Durbin-Watson stat0.647473Pro b(F-sta.ti Stic)0.000000图2(2) 模型的结果见图2,可以得出y、xl、x2、都是逐年增长的，但增长率有所变 动，说明变量间不一定是线性关系，可以设定模型为：Y=69.07x1+6.68x2+(-24628.71)(2)S=l8.l50.25l26l0.57T= 3.8026.49-l.95R 2 =0.964 R2 =0.962 F=364.1973P=0(3) 置信区间在a=0.05显著性水平上，查自由度为n-k-l=30-2-l=27的t分布 表，得临界值ta/2(n k-1)二爲匚2.0518，所以，置信度为95%的bi 的置信区间为/ AAAA(b t (27) s(b),b +1(27) s(b )10.025110.0251=(69.07-2.05*l8.l5 ， 69.07+2.05*l8.l5)=(31.86，106.28)同理，置信区间为 95%的 b 2的置信区间为/ AAAA(b t (27) s(b ),b +1 (27) s(b )2 0.025 2 2 0.025 2=(6.68-2.05* 0.25，6.68+2.05* 0.25)=(6.17,7.19)(三)模型检验(1) 、经济意义检验：1、回归系数估计值t? =69.070, = 6.680,说明汇率和外汇储1 2备的变化对进出口总额与汇率和外汇储备指数同方向变动,当其他条 件不变时,汇率和外汇储备每增加一个百分点,进出口总额将依次增 加69.07%、 6.68%。2、回归方程的标准误差评价：S.E=17517.93,说明进出口总额指数观测点与其估计值的平均误差为 17517.93。(2) 、统计意义检验：1、拟合优度检验：R2 = 0.9616说明，回归方程的解释能力为96.16%,即汇率和外汇储备对进出口总额的解释能力为 96.16%,回归方程的拟合优度非常好。这从残图图3也可以看出来。0File Edit Object View Proc Quick Options Add-ins Window Help data, y x1 x2Is y cx1 x2H Workfile: UNTITLED Group: UNTITLED Workfile: UNTITLED:UntitledView Pro2066.7-300,000Resids1985198619871988198919901991199219931994199519961997| View|procObject| print|Name |Freeze | Estimate |Forecast |Stats目 Equation: UNTITLED Workfile: UNTITLED:Untitled199819992000200120022003200420052006-50,000-50.000-250,000-200,000-150,000-100,000图32、回归模型的总体显著性检验：从全部因素的共同影响看，在 5%的显著水平上,F=364.1973(k,n-k-l) = (2,27)=3.35,说明汇率和 外汇储备对进出口总额的共同影响是显著的。这从P=0,可以看出回 归模型是显著的。3、单个回归系数的显著检验：从单个因素的影响看,在5%的显 著水平上，t(X )二3.804837(27) =2.0518,说明说明汇率对进出口1总额的影响是明显的，t(X ) = 29 49368(27) =2.0518，说明外汇储2备进出口总额影响也是显著的，这从他们的 p 值为 0.0007、0.0000 也可以看出回归系数显著不为零。(3) .计量经济学检验:1、多重共线性检验:图4由相关系数矩阵可以看出，各解释变量相互之间的相关系数较 高，解释变量之间确实存在严重多重共线性。图5Y=6.73x2+20962.87 (3)S=4783.246 0.306582T=4.382561 1.95334R 2 =0.945 后=0.943 F=481.95P=0 2、异方差检验:0File Edit Object View Proc Quick Options Add-ins Window Help Is y cx1 x2coryxl x2Isyx2View ProH Workfile: UNTITLED=Equation: UNTITLED Workfile: UNTITLIED:Untitled - n x| View|(Proc| Object printName |Freeze | | Estimate |Forecast |Stats |Res idsF-statistic0.069033Prob. F(1,28)0.7947Obs*R-squared0.073782Prob. Chi-Square(l)0.7859Scaled explained! SS0.065841Prob. Chi-Square(l)0.7975Heteroskeelasticity Test: WhiteSaIASM0MTest Equation:Dependent Variable: RESIDA2 Method: Least SquaresDate: 01/06/16 Time: 15:06Sample: 1985 2014 Included observations: 30VariableCoefficientStd. Errort-StatisticProb.C7.42E+032.13E+083.4909230.0016X2A2-0.1099140.418337-0.2627410.7947R-squared0.002459Mean dependent var7.15E+08Adjusted R-squared-0.033167S.D. dependent var1.01E+09S.E. of regression1.02E+09Akaike info criterion4439229Sum squared resid2.92E+19Schwarz criterion44.48571Log likelihood-663.8844Hannan-Quinn criter.44.42218F-sta.tistic0.069033Durbin-Watson stat0.239602Pro b(F-stati stic)0.794674图6P0.05不存在异方差。3、自相关检验:=J tquation: UNIIILtD Worktile: UNIIILtD:Untitled xView| Proc| Object printName I Freeze| (Estimate | Forecast | Stats | Res idsBreusch-Godfrey Serial Correlation LM Test:AF-sta.tistic20.70915Prob. F(2,26)0.0000Obs*R-squared18.43044Prob. Chi-Square(2)0.0001Test Equation:Dependent Variable: RESIDMethod: Least SquaresDate: 01/06/16 Time: 15:07Sample: 19852014Included observations: 30Presample missing value lagged residuals set to zero.VariableCoefficientStd. Errort-Sta.tisticProb.C822.91123121.7070.2636090.7942X2-0.1233460.210979-0.5846370.5638RESIDf-1)0.8927150.1948894.5806400.0001RESID(-2)-0 1355320.206779 0.6554410.5179R-squared0.614348Mean dependent var-7.28E-12Adjusted! R-squa.red0.569850S.D. dependent var20950.11S.E. of regression1374030Akaike info criterion22.01762Sum squared resid4.91 E+09Schwarz criterion2220445Log likelihood-326.2643Hannan-Quinn criter.22.07739F-statistic13.80610Durbin-Watson stat1.897979V图7P0.05存在自相关。消除自相关：File Edit Object View Proc Quick Options Add-ins Window HelpIsyx2 ar Is y c x2H Workfile: UNTITLED | ViewI Proc | Object IprintjSaIA0M0MY1.0(X10.1EX20.97I View I Proc Object | | Print I Save | Details+/-j Show I Fetch I Store I Delete | Genr | Sample |Range: 1985 2015 一 31 obsSample: 1985 2015 一 31 obssi(eCresx1x2yyfyfsFilter:*回 Equation: UNTITLED Workfile: UNTITLED:Untitled - n xI View I Proc I Object I Print I Name I Freeze I Estimate | Forecast I Stats | Resids I Untitled New Page |VF -P:S-E. |Forecast: YFAC1I3I： Yforecast sample 1935 2015Aapaed sample 1986 2015 nciudMooservaeons 29Wt Man soareo Error1928309k*eanAEouee Error1632225leanADB Peroer* Error97 9SM1Tneii nequawy CoencJert0 08097SProportion0.003735Vartrve Proportion0290115Couartance Proportion0706150图12hFile Edit Object View Proc Quick Options Add-ins Window HelpIs y c x2Is y c x2 ar(1)/0 Series: YF Workfile: UNTITLED:Untitled- BX田 worKTiie; uiView Proc ObjectView| Proc Object | Properties | Print | Name | Freeze | Defaultv | Sort Edit*/-smpHYFRange: 1985 201sample: iy5 2Ui19867903.564AVc0 resid0x10x20y0yf198712903.73198817106.77198920755.08199024043.94199127156.940 yfse199229143.91199331029.19199434267.47199536789.46199639651.14199742540.45199843592.03199944788.91200045952.52200149063.47200253677.21200360656.22200472699.46200584828.32200699114.632007125580.32008149507.62009175425.42010201043.72011220134.12012227631.92013256714.62014258000.620152309615.图130File Edit Object View Proc Quick Options Add-ins Window HelpIs y c x2Is y c x2 ar(1)0 Series: YFSE Workfile: UNTITLE D:Untitled nXView Proc | Objectl Properties | Printl Name| Freeze | Defaultv | Sortl Edit+/-|smphYFSEViewPIRange:SampleA1985NA1 cM re198613661.93si198717803.230x1198820215.250x2198921765.970y199022S12.08y yrM yfse19912353835199224072.4419932446139199424719.75199524922.74199625065.99199725170.74199825287.67199925374.28200025441.59200125449.62200225422.16200325356.76 U200425252.92200525172.49200625117.48200725280.72200825435.74200925720.26201026058.48201126200.0320122604432201326S29.7420142648439I1-2015402512.0M Seriv图14三、结论回归系数估计值，S二-0.1680,说明外汇储备的变化对进出口2总额反方向变动，当其他条件不变时，汇率和外汇储备每增加一个百 分点，进出口总额将依次减少0 . 1 6 8%。