VAR模型應(yīng)用案例-

VAR模^(moxing)應(yīng)用案例-(完成)VAR模型應(yīng)用(yingybng)案例-(完成)

VAR模型（mdxlng）應(yīng)用實(shí)例眾所周知，經(jīng)濟(jì)的發(fā)展運(yùn)行離不開大量能源的消耗，尤其是在現(xiàn)代經(jīng)濟(jì)發(fā)展的過程中，能源的重要（zhbngydo）性日益提升。我國自改革開放以來，經(jīng)濟(jì)發(fā)展取得長足的進(jìn)步，經(jīng)濟(jì)增長率一直處于較高的速度，經(jīng)濟(jì)的高速增長帶來了能源的大量消耗，進(jìn)而帶來了我國能源生產(chǎn)的巨大提高。因此，研究經(jīng)濟(jì)增長率與能源生產(chǎn)增長率之間的關(guān)系具有重要的意義，能為生源生產(chǎn)提供一定的指導(dǎo)意義。1基本（jib&n）的數(shù)據(jù).我們（wdmen）截取1978一2015年中國（zhonggud）經(jīng)濟(jì)增長速度（GDP增速）和中國能源生產(chǎn)增長速度數(shù)據(jù)，具體數(shù)據(jù)如下：表11978——2016年中國經(jīng)濟(jì)和能源生產(chǎn)增長率年份國內(nèi)生產(chǎn)總值增長速度（%）能源生產(chǎn)增長速度（%）年份國內(nèi)生產(chǎn)總值增長速度（%）能源生產(chǎn)增長速度（%）197811.710.419797.63.719807.81.319815.10.8198295.6198310.86.7198415.29.2年份國內(nèi)生產(chǎn)總值增長速度（%）能源生產(chǎn)增長速度（%）19979.20.319987.82.719997.71.620008.5520018.36.420029.1620031014.1198513.49.9200410.115.619868.93200511.411.1198711.73.6200612.76.9198811.25200714.27.919894.26.120089.7519903.92.220099.43.119919.30.9201010.69.1199214.22.320119.59199313.93.620127.93.21994136.920137.82.21995118.720147.30.919969.93.120156.91.22序列(xdli倉)平穩(wěn)性檢驗(yàn)(單位根檢驗(yàn)).使用(shiydng)Eviews9.0來創(chuàng)建(chuangjian)一個(gè)無約束的VAR模型(mdxlng),用gdp表示的是中國(zhOnggu6(mdxlng),用gdp表示的是中國(zhOnggu6)經(jīng)濟(jì)的增長率，用nysc表示中國能源生產(chǎn)的增長=i=i率，下面分別對gdp和nysc進(jìn)行單位根檢驗(yàn),驗(yàn)證序列是否平穩(wěn)，能否達(dá)到建立VAR模型的建模前提。

MSeries:GDPWorkfile:UNTITLED::Untitled\三.」!旦1ViewProcObjectPropertiesPrintNameFreezfSampleGemSheet|Graph!AugmentedDickey-FullerUnkRootTestonGDPNullHypothesis:GDPhasaunitrootExogenous:ConstantLagLength:3(Automatic-basedonSIC,maxiag=9)t-StatisticProb.1Augmen悅dDick時(shí)-RjIleftiststatistic -3867553 。如刷Testcriticalvalues: 1%level -3 6394075%level -2 95112510%level -2 614300,LlacKjnnon(1996)one-sidedp-values.AugmentedDickey-FullerTestEquationDependentVariable:D(GDF)Method:LeastSquaresDate:05/17/17Time:10.55Sample(adjusted):19822015Includedobservations:34afteradjustmentsVariableCo&ffidentStdErrort-StatisticProbGDP(-1)-0.8551710.2211143867553Q.DC06DfGDP{-1})0.6256310.1935293.2327550.0031D{GDP(-2)J0.0492400.17551702805440.7G11D[GDP(-3))0.264S370.1673481.5831450.1242C&.5400502222961384174500006R-sqjared 0.45B475Meandependentvar 0.052941AdjustedR-squared 0.3S37B2SD.dependentvar 2.545731圖2.1經(jīng)濟(jì)(jingji)增速(GDP)的單位根檢驗(yàn)(ji^nydn)SSerieE!N￥SCWorkfik:UNTITLED”Untitled、 !o*dViewProcObjectPropertiesPrintHflmeFreeze.SampleGenrSheetGrciphAugmentedDickey-FullerUnitRootTestonNYSCMl|UH/pottiesis:NYSChmsaunitrootExogenous:ConstantLagLength:1(Automatic-basedonSIC,maxlag=9)t-StatistcProb?Augment日dD無席y-FuI便「怕ststatistic -39359$? 。一。。45Testcriticalvalues: 1%level -16257845%level -2.94584210%level -2.611531TLl3cKinnon(1996)one-sidedp-vaiuesAugmentedDickey-FullerT盹tEquationDependentVariaWe:D(rJYSC)Melhod:LeastSquaresDale:Qb/17?17Tme:10:53Sample(adjusted):19302015indudedobserations:36afteradjustmentsVariableCoefficientStd.Errort-StatisticProsNYSC(-1)-0.5309S60134905 -39359870.0004D(NYSC(-1))04385490150055 29225B500062C2.7469380857266 32043000.0030R-squared0.343068Meandependentvar-0D6S444AdjustedR-squared0.3D3254S.D.dependentvar3510704S.E.ofregression2.930431Aksifceinfoenterion5D67B31SumgquM&dr&sid283.3351Sctiwarzcriterion5193791Logli暗由。時(shí)-88.22Q9eHannan-Quinncriter5.11擁四F-statistk8.81B746DuFbin-WatstHistat1.99D251PramF-stetistic}0.000975圖2.2能源(nSngyudn)生產(chǎn)增速(nysc)的單

位根檢驗(yàn)(jiSnyan)經(jīng)過(jlnggub)檢驗(yàn)，在1%的顯著性水平(shuiping)上,gdp和nysc兩個(gè)(lidngg《)時(shí)間序列都是平穩(wěn)的，符合建模的條件，我們建立一個(gè)無約束的VAR模型。3VAR模型的估計(jì).ViewProcObjectPrintNairitFreezeEstimateForecast5VectorAuioregressionEsttmatesVector.AutoregressionEstimatesDate:05/17/17Time:11:03Sample(adjusted):19802015Includedobsenations:36afteradjustmentsStandarderrorsin()At-stallstrcsin[]GOPNYSCGDP^-1)0.925544027159S(0.16499)(023599)(5.00369][115086JGDP{-2)-0.530495-0.292356[0.16525)(0.23700)1-3.19096][-1.22942]NYSC(-1)-0.0522250.34635&(0.11565)(0,16543)1-0.45156][511612]N￥SC(-2)0186100■0.3575舶(0.11343)(0.16234)[163977][-2.20263JC6.19451S23G3291(1.50867)(2.15827)(4.10539][1.32666iR-squared0.4925650.5&4337Ad]R-squared0.4270390,4963&9Sumsq.resids130.5151257X333SEequation2.0518692934958F'Statistic7.S22S909.641751Laglikelihood-74.26525AKalICBAIC4.4036255.119509Schwarzsc4.62355S5.339442Meandependent97388895.016667S.D.dependsnl2.71035^4.137805Determinantresidccvariancefootacfj.J30.72390Determinantresidcovariance23.7B21&Lagli^elihciod-1584312AkaiKeinformationcriterion93572S7Schwarzcriterian9.797154圖3.1模型(mdxlng)的估計(jì)結(jié)果回Viir;VAR01WQfkfile;制描WAR槿型;;Untitled^ViewProcOtyedPrintIJjme[Freeze!Estima-teForecatiStatsImpureResidsitiB&tiKnFree：■LS12GBF1RSC?JlodtJ:;（?DP=CCLlhGBiPf-l）*「（1,團(tuán)"耽-2）+CtL3）mSC（-l）*C（L4）*MTfSC（-2j'c（LE）1TYSC=C^l^GWPt-l）*C（2h2）^BP（-2）+C（2,3）*R?SC（-1）/C（A4）*跑既：（-2）7AHMed成一 ents;<?DF=Q.S255W312fl35*GDFH）-U.5304S4707464^EF（-￡）-□.OKE24TO551Q2tNYSCH）+8一1蹄100頌花4栩TSC（Y）+6.194Sl&W4f&31TYSC=02?t55I3i9?Sr4*（；DFH）-0.?3Z35fiia6154*GDr（-?）+0.846355866747?NYSCH）-0.36?56?63￡7*6^30（^）十2.8S3291OBL70圖3.2模型（mdxing）的表達(dá)式4.模型(mdxlng)的檢驗(yàn)4.1模型(m6xlng)的平穩(wěn)性檢驗(yàn)回Var:UNTITLEDWorkfilt:UNTITLED::UntftledXr~T tt1r=^T-11 T"ViewProcObjectPrintHame:FreezeEstimateForecastSI VARStafatttyCftiwmiMiOwcfcRootsofCharacteriSticPolynomialEndogenousvariables;GDPNYSCExogenousvariables:CLagspecification:12Date;05/17/17Time;11;11Root Modulus0.566036-0.4517081 0.7242200.566086+0.451708i &.7242200769064-0.6.265511 tt.6821960.269864+0.6?6551l 0.682196Norootliesoutsidetheunitcircle.VARsatisfiesthestaNlit^'candition.圖4.1.1AR根的表

由圖4.1.1知，AR所有(su6y6u)單位根的模都是小于1的，因此(yinci)估計(jì)的模型滿足穩(wěn)定性的條件。I-rnrw]vw■■■i-^ipw iimus.i?wv-i一％ui■■iipwrrir—IrvAFAAofARCharact^rEAttrPohynomiaI圖4.1.2AR根的圖通過(tongguo)對GDP增長率和能源(n^ngyudn)生產(chǎn)增長率進(jìn)進(jìn)行(jinxing)7VAR模型估計(jì)，并采用AR根估計(jì)的方法對VAR模型估計(jì)的結(jié)果進(jìn)行平穩(wěn)性檢驗(yàn)。AR根估計(jì)是基于這樣一種原理的:如果VAR模型所有根模的倒數(shù)都小于1,即都在單位圓內(nèi)，則該模型是穩(wěn)定的;如果VAR模型所有根模的倒數(shù)都大于1，即

都在單位圓外，則該模型是不穩(wěn)定的。由圖4.1.2可知，沒有根是在單位圓之外的，估計(jì)的VAR模型滿足穩(wěn)定性的條件。4.2Granger因果(ylngu6)檢驗(yàn)圖4.2.1Granger因果檢驗(yàn)(jiEnydn)結(jié)果圖Granger因果(yingub)檢驗(yàn)的原假設(shè)(j訝sh?)是：H:變量(bidnlidng)x不能Granger引起變量0y備擇假設(shè)是：H1：變量x能Granger引起變量y對VAR(2)進(jìn)行Granger因果檢驗(yàn)在1%的顯著性水平之下，經(jīng)濟(jì)增速(GDP)能夠

=1Granger引起能源生產(chǎn)增速(NYSC)的變化，即拒絕了原假設(shè)；同時(shí)，能源生產(chǎn)增速=1(NYSC)能夠Granger經(jīng)濟(jì)增速(GDP)的變化，即拒絕了原假設(shè)，接受備擇假設(shè)。5滯后(zhihdu)期長度s'VanUNTITLED UNT(TLEC)：；Urttitl?d\1'/i?w|ProcjObject||PrintjNome|Frc?e11Estimate|ForecastStrts]lmpulsejResidj|ZcomVARLagOrderSelectionCriteriaEndosen&usvariables:GDPNYSCExogenous CDate-05^17/17Time11.16Sample:19783015Injucfedobservaticns34LagLogLLRFPEAJCSCHQ0-172.7423NA99.S069610.27S961036B7410.3095S1-158.355025.323&455.841d99.59T3519.9567099.7692102-148.079713.38127*37.61SQ3*9.29胡。B*S.7477S8*J.4519O&*3-147,4983D.92353446.3436594998971O.12B40"142344-145.49942.94091S52.858889.51755510.42S6J拙93131■mcticateslagorderselectedD<tnecntenonLR：sequenlislTiodiTiedLRleststatisllc(eachtest3t=%level)FPE：FinalpredictionerrorAIC：informationcriterionSC:SchwarzmiormationcriterionHQ:Hannan-Quinninlornationcriterion圖5.1VAR模型滯后(zhihdu)期選擇結(jié)果

從上圖可以(k^yl)看出LR,FPE,AIC,SC,HQ都指向(zhixiang)同樣的2階滯后期，因此(yinci)應(yīng)該選擇VAR(2)進(jìn)行后續(xù)的分析。

6.脈沖(mdichong)函數(shù)回Van￥AR01Workfiie;殼隘的VAR槿型;;FreezeEsticci■日teF。比Stab4mpuhsResideR&3KF19St-GJw^skyOwG.D.IrnoT-WsT2W_E回Van￥AR01Workfiie;殼隘的VAR槿型;;FreezeEsticci■日teF。比Stab4mpuhsResideR&3KF19St-GJw^skyOwG.D.IrnoT-WsT2W_EView.ProcOtye-rtPrintNameRt^juM8iJF<VSGoGDPArapononarmraG^NYSG經(jīng)濟(jì)(jingji)增長率(GDP)和能源(n^ngyudn)生產(chǎn)(NYSC)各自對于自身的沖擊，在前四期是快速下降的趨勢，并且出現(xiàn)負(fù)

值的情況。但是，GDP增速的變化基本上在第七期就保持了持平的一個(gè)狀況；而能源生產(chǎn)(NYSC)的變化是在第九期的時(shí)候?qū)崿F(xiàn)持平的狀態(tài)。=1=1=1能源(nOngyudn)生產(chǎn)增長率(NYSC)對于(duiyii)經(jīng)濟(jì)增長率(GDP)的脈沖響應(yīng)分析，當(dāng)給經(jīng)濟(jì)增長一個(gè)(yigd)正的沖擊的時(shí)候，在前兩期是呈現(xiàn)一個(gè)下降的趨勢，主要的原因(yudnyin)應(yīng)該是，經(jīng)濟(jì)增長促進(jìn)能源生產(chǎn)的提高是存在滯后期的，但是(ddnshi)但很快就出現(xiàn)了上升的趨勢在第五期的時(shí)候達(dá)到最大值，之后出現(xiàn)了下降的趨勢，然后又回升，直到第十期之后保持了平衡。這說明經(jīng)濟(jì)增長對于能源生產(chǎn)增長的影響是正向的，會(huì)呈現(xiàn)一種上升、下降、平衡的基本狀態(tài)，說明經(jīng)濟(jì)發(fā)展對能源生產(chǎn)的促進(jìn)作用并不是無限的，經(jīng)過一定作用之后看，會(huì)出現(xiàn)一種平衡狀態(tài)。=1=1=1經(jīng)濟(jì)增長率(GDP)對于能源生產(chǎn)增長率(NYSC)的脈沖響應(yīng)分析，經(jīng)過對比圖中第2幅和第3幅小圖，我們大致是可以看出兩者之間是呈現(xiàn)完全相反的情況。當(dāng)在本期給能源生產(chǎn)增長率(NYSC)一個(gè)正沖擊之后，前兩期是

增長，然后到第五期是下降趨勢，然后回升，在第七期之后基本上持平。7.方差分析^3VanUNTITLEDWorkfile:UNTfTLED::UntMed\ViewPkcObje(i|PrintNameFreezeEstimateForecastStatsImpureR芯i也ZoomVarianceDecompositi<uiVsriance□ecompositionorGDP:APeriodSEGDPNYSC12.051&69100.00^00.00000022.S2678299.711540.2蹄45832669&5390.721431.2785704276860692919477.08053352B4515389.1101110.SB&89&2.851171808871311.1128772.B588273S.430A71156&93S2.07241087.B373412.3626692.S7677797.4049012.59510102.076^0S7402311259769112B775-48873871112.61289122.878296873730712.62&9313267848187.3670712.63293142.873&2587.365。912.63491152.87857887,3550812.鏘492162.B7S&01B73652412.63476172.B78&1387364洗12.63534182.078625873639212.63&08192.873629873636S12.63B32202.S7S&3087.3636912.C3&31212B7S&31873636412.63&36222.B7S&32S73635812.63642232.87863387.3635612.63644242.878&33873635&12.63644252.878&3387.3635€12.C3644262.B78&33873635612.63644272.B78&33S73635612.63&44282.87明3307.3635612.63544292878&3387.3B35612.63&44302.87B&3387.3G35612.63644 J圖7.1經(jīng)濟(jì)（jingji）增長（GDP）方差分析結(jié)果（ji。gu6）

回V,anUNTITLEDWarlcfile:UNTrLED::Untitled\5回心TiewPfo-cObjectPrintCarrieFreezeIEHim日坨StmtsFmpulseR閔。5ZoomVarianceDecom[>ositionL?arianceDecompositionofNVSC:APeriodSEGDPNYSC12.93.495315.28242S47175824.02209122.57C7677.42324341?1943237934B762065244.20-352324.1765175.8234954,25518425.247B573.7520564.2&887527.145707285430743&702327.099蹈729007284J14780270076472.93236g431S40127JQ6767289324104.31S99627.125227287478nf432022527.1145572885451；4.321254271025772.&0743i4.321463270999472.9000