實(shí)驗(yàn)七 滯后效應(yīng)、虛擬變量、時(shí)間序列和聯(lián)立方程模型的估計(jì)-學(xué)生實(shí)驗(yàn)報(bào)告

廣東財(cái)建大辱辜商孕憂HUASHANGCOLLEGEGUANGDONGUNIVERSITYOFFINANCE&ECONOMICS實(shí)驗(yàn)報(bào)告課程名稱：計(jì)量經(jīng)濟(jì)學(xué)實(shí)驗(yàn)項(xiàng)目實(shí)驗(yàn)七滯后效應(yīng)、虛擬變量、時(shí)間序列和聯(lián)立方程模型的估計(jì)實(shí)驗(yàn)類型:綜合性口設(shè)計(jì)性口驗(yàn)證性專業(yè)班別：姓名：學(xué)號(hào)：實(shí)驗(yàn)課室：厚德樓A404指導(dǎo)教師：實(shí)驗(yàn)日期：2015年6月25日星期四廣東商學(xué)院華商學(xué)院教務(wù)處制、實(shí)驗(yàn)項(xiàng)目訓(xùn)練方案小組合作：是口否小組成員：無實(shí)驗(yàn)?zāi)康模赫莆諟笮?yīng)、虛擬變量、時(shí)間序列和聯(lián)立方程模型的估計(jì)實(shí)驗(yàn)場(chǎng)地及儀器、設(shè)備和材料實(shí)驗(yàn)室：普通配置的計(jì)算機(jī)，Eviews軟件及常用辦公軟件。實(shí)驗(yàn)訓(xùn)練內(nèi)容（包括實(shí)驗(yàn)原理和操作步驟）：【實(shí)驗(yàn)原理】分布滯后模型、自回歸模型虛擬解釋變量模型單位根檢驗(yàn)、協(xié)整和誤差修正模型聯(lián)立方程模型。【實(shí)驗(yàn)步驟】（一）滯后變量模型1、分布滯后模型（課本P178）仿照課本例7.5（全部?jī)?nèi)容），建立模型，分析我國(guó)居民消費(fèi)價(jià)格TBZS受貨幣供應(yīng)增長(zhǎng)量M2Z影響的模型。數(shù)據(jù)見“全國(guó)廣義貨幣供應(yīng)量及物價(jià)指數(shù)月度數(shù)據(jù)”。

DependentVariable:TBZSMethod:LeastSquaresDate:06/25/15TimA:10:49SamplA(adjusted):1996M022008M11Includedobservations:154afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.C101.36930.347947291.33520.0000M2Z0.2958730.0994442.9752620.0034R-squared0.055033Meandependentvar102.1383AdjustedR-squared0.048816S.D.dependentvar2.964169S.E.ofregression2.890915Akaikeinfocriterion4.973925Sumsquaredresid1270.323Schwarzcriterion5.013366Loglikelihood-380.9922Hannan-Quinncriter.4.989946F-statistic8.852184Durbin-Watsonstat0.144830Prob(F-statistic)0.003406從回歸結(jié)果看，M2Z的t統(tǒng)計(jì)量顯著，表明當(dāng)期貨幣供應(yīng)量的變化對(duì)當(dāng)期物價(jià)水平有一定影響但沒有顯現(xiàn)出這種影響的滯后性。為了分析貨幣供應(yīng)量變化影響物價(jià)的滯后性，我們作之后6個(gè)月的分布滯后模型的估計(jì)，在EVIEWS工作文檔的方程設(shè)定窗口中,輸入TBZSCM2Z(-1)M2Z(-2)M2Z(-3)M2Z(-4)M2Z(-5)M2Z(-6)結(jié)果見表DependentVariable:TBZSMethod:LeastSquaresDate:06/25/15Time:10:52Sample(adjusted):1996M082008M11Includedobservations:148afteradjustmentsVariableCoefficientStdAErrort-StatisticProb.C99.226610.386141256.97020.0000M2Z0.0477670.0964260.4953700.6211M2Z(-1)0.1340200.0915651.4636680.1455M2Z(-2)0.1573680.0904571.7396960.0841M2Z(-3)0.1521170.0927761.6396200.1033M2Z(-4)0.1799260.0901571.9957000.0479M2Z(-5)0.1666960.0922531.8069390.0729M2Z(-6)0.1799740.0971701.8521580.0661R-squared0.305349Meandependentvar101.8561AdjustedR-squared0.270617S.D.dependentvar2.659733S.E.ofregression2.271517Akaikeinfocriterion4.531311Sumsquaredresid722.3706Schwarzcriterion4.693323

LoglikelihoodF-statisticProb(F-statistic)LoglikelihoodF-statisticProb(F-statistic)4.5971360.0959978.791439Durbin-Watsonstat0.000000從回歸結(jié)果看，M2Z各滯后期的系數(shù)逐漸增加，表明當(dāng)其貨幣供應(yīng)量的變化對(duì)物價(jià)水平的影響要經(jīng)過一段時(shí)間才能逐步顯現(xiàn)。但各滯后期的系數(shù)的t統(tǒng)計(jì)量值不顯著，因此還不能據(jù)此判斷滯后期究竟有多長(zhǎng)。為此，我們作滯后12個(gè)月的分布滯后模型的估計(jì)，結(jié)果如下：DependentVariable:TBZSMethod:LeastSquaresDate:06/25/15Time:10:53Sample(adjusted):1997M022008M11Includedobservations:142aftAradjustmentsVariableCoefficientStd.Errort-StatisticProb.C98.199750.313325313.41220.0000M2Z-0.0649220.086361-0.7517470.4536M2Z(-1)0.0795070.0784611.0133300.3128M2Z(-2)0.0686290.0816670.8403530.4023M2Z(-3)0.0995560.0822801.2099690.2285M2Z(-4)0.1324290.0828811.5978280.1125M2Z(-5)0.0442900.0822150.5387140.5910M2Z(-6)0.0678940.0821240.8267220.4099M2Z(-7)0.1316240.0822361.6005620.1119M2Z(-8)0.1526020.0824871.8500020.0666M2Z(-9)0.0854950.0822461.0395020.3005M2Z(-10)0.0782950.0814440.9613310.3382M2Z(-11)0.2047460.0948262.1591700.0327M2Z(-12)0.2889870.1007072.8695750.0048R-squared0.554030Meandependentvar101.6366AdjustedR-squared0.508737S.D.dependentvar2.482034S.E.ofregression1.739662Akaikeinfocriterion4.038645Sumsquaredresid387.3823Schwarzcriterion4.330065Loglikelihood-272.7438Hannan-Quinncriter.4.157066F-statistic12.23193Durbin-Watsonstat0.201551Prob(F-statistic)0.000000通過上述一系列分析，我們可以做出這樣的判斷：在我國(guó)，貨幣供應(yīng)量變化對(duì)物價(jià)水平的影響具有明顯的滯后性,滯后期大約為四個(gè)季度，而且滯后影響具有持續(xù)性，持續(xù)的長(zhǎng)度大約為半年左右，其影響力度先遞增后遞減，之后結(jié)構(gòu)為人型。根據(jù)前面的分析可知，分布滯后模型可以用自回歸分析滯后模型來代替，因此我們估計(jì)如下自回歸模型:TBZS=a+01*TBZS(-1)+02M2Z+u

dependentVariable:TBZSMethod:LeastSquaresDate:06/25/15Time:10:54Sample(adjusted):1997M082008M11Includedobservations:136afteradjustmentsVariableCoefficientStd.Errort-StatisticProb.C97.596530.286256340.94150.0000M2Z-0.0195200.077185-0.2529000.8008M2Z(-1)0.0150640.0772860.1949140.8458M2Z(-2)-0.0205390.079295-0.2590190.7961M2Z(-3)0.0043090.0790560.0545060.9566M2Z(-4)0.0015230.0812150.0187520.9851M2Z(-5)0.0047860.0824890.0580230.9538M2Z(-6)-0.0117630.081670-0.1440270.8857M2Z(-7)0.0669610.0787200.8506160.3967M2Z(-8)0.0917570.0783921.1704860.2442M2Z(-9)0.0431190.0783850.5500860.5833M2Z(-10)0.0364990.0773710.4717400.6380M2Z(-11)0.1645430.0870291.8906690.0612M2Z(-12)0.2142240.0948302.2590270.0257M2Z(-13)0.2317050.0944852.4522830.0157M2Z(-14)0.2124500.0956592.2209050.0283M2Z(-15)0.2154320.0970112.2206870.0283M2Z(-16)0.1721570.0961301.7908780.0759M2Z(-17)0.1094690.0968741.1300150.2608M2Z(-18)0.1148720.0920971.2472980.2148R-squared0.685178Meandependentvar101.5537AdjustedR-squared0.633612S.D.dependentvar2.494614S.E.ofregression1.509989Akaikeinfocriterion3.797135Sumsquaredresid264.4877Schwarzcriterion4.225466Loglikelihood-238.2052Hannan-Quinncriter.3.971198F-statistic13.28748Durbin-Watsonstat0.197989Prob(F-statistic)0.0000002、自回歸模型（實(shí)驗(yàn)指導(dǎo)書p180）根據(jù)數(shù)據(jù)，建立廣東省城鄉(xiāng)儲(chǔ)蓄存款CX的自回歸模型（作一階自回歸模型、考慮LB、RK作為自變量）。并解釋模型的實(shí)際意義。數(shù)據(jù)見“廣東省宏觀經(jīng)濟(jì)數(shù)據(jù)-實(shí)驗(yàn)七”。DependentVariable:CXMethod:LeastSquaresDate:06/25/15Time:10:59Sample(adjusted):19792005Includedobservations:27afteradjustmAntsVariableCoefficientStd.Errort-StatisticProb.CX(-1)0.7478250.0850618.7916340.0000LB0.9840970.2173404.5279170.0002RK-0.6880020.237546-2.8962930.0081C3588.7051289.7822.7824120.0106R-squared0.998181Meandependentvar4446.959AdjustedR-squared0.997944S.D.dependentvar5613.364S.E.ofregression254.5348Akaikeinfocriterion14.05271Sumsquaredresid1490123.Schwarzcriterion14.24468Loglikelihood-185.7115Hannan-Quinncriter.14.10979F-statistic4207.399Durbin-Watsonstat1.017355Prob(F-statistic)0.000000CX(-1)不僅顯著，LB、RK也顯著，方程為CX=0.747825*CX(-1)+0.984097*LB-0.688002*RK+3588.7047（請(qǐng)對(duì)得到的圖表進(jìn)行處理，以上在一頁(yè)內(nèi)）（二）虛擬解釋變量模型（請(qǐng)對(duì)得到的圖表進(jìn)行處理，以上在一頁(yè)內(nèi)）（二）虛擬解釋變量模型（實(shí)驗(yàn)指導(dǎo)書P200）根據(jù)數(shù)據(jù)，考慮1994年的稅制改革，作為虛擬變量引入到稅收CS對(duì)生產(chǎn)稅SE的模型中，建立合理的模型。數(shù)據(jù)見“廣東省宏觀經(jīng)濟(jì)數(shù)據(jù)-實(shí)驗(yàn)七”。

ResidualActualFittedDependentVariable:CSMethod:LeastSquaresDate:06/25/15Time:11:05Sample:19782005IncludedobservationsA28VariableCoefficientStd.Errort-StatisticProb.SE0.6211710.0A500641.394130.0000C20.527359.7832572.0982120.0462DD94-139.593527.13496-5.1444160.0000R-squared0.994770Meandependentvar449.5546AdjustedR-squared0.994352S.D.dependentvar509.5465S.E.ofregression38.29494Akaikeinfocriterion10.22947Sumsquaredresid36662.57Schwarzcriterion10.37221Loglikelihood-140.2126Hannan-Quinncriter.10.27311F-statistic2377.613Durbin-Watsonstat2.197491Prob(F-statistic)0.000000

ResidualActualFittedDependentVariable:CSMethod:LeastSquaresDate:06/25/15Time:11:06Sample:19782005Includedobservations:28VariableCoefficientStd.Errort-StatisticProb.SE0.5556810.01128149.259170.0000C14.6064713.621151.0723380.2938D94-66.2242254.39562-1.2174550.2348R-squared0.989836Meandependentvar449.5546AdjustedR-squared0.989023S.D.dependentvar509.5465S.E.ofregression53.38516Akaikeinfocriterion10.89390Sumsquaredresid71249.37Schwarzcriterion11.03664Loglikelihood-149.5146Hannan-Quinncriter.10.93754F-statistic1217.373Durbin-Watsonstat1.306956Prob(F-statistic)0.000000建立的財(cái)政收入CS的回歸方程為：CS=0.556143052611*SE+11.8774094587

（請(qǐng)對(duì)得到的圖表進(jìn)行處理，以上在一頁(yè)內(nèi)）課后拓展以下第三部分和第四部分屬于課后自學(xué)、拓展練習(xí)。（三）時(shí)間序列模型（實(shí)驗(yàn)指導(dǎo)書P182）根據(jù)數(shù)據(jù)，建立廣東省城鎮(zhèn)居民的人居可支配收入RJSR與人均消費(fèi)水平RJXF的模型。數(shù)據(jù)見“廣東省宏觀經(jīng)濟(jì)數(shù)據(jù)-實(shí)驗(yàn)七”。1、單位根檢驗(yàn)對(duì)以上模型的數(shù)據(jù)進(jìn)行單位根檢驗(yàn)。2、協(xié)整檢驗(yàn)和誤差修正模型對(duì)以上模型作協(xié)整檢驗(yàn)和建立誤差修正模型。）聯(lián)立方程模型（實(shí)驗(yàn)指導(dǎo)書P220）用二階段最小二乘法估計(jì)財(cái)政支出方程。數(shù)據(jù)見“廣東省宏觀經(jīng)濟(jì)數(shù)據(jù)-實(shí)驗(yàn)七”。二、實(shí)驗(yàn)總結(jié)與評(píng)價(jià)實(shí)驗(yàn)總結(jié)（包括實(shí)驗(yàn)數(shù)據(jù)分析、實(shí)驗(yàn)結(jié)果、實(shí)驗(yàn)過程中出現(xiàn)的問題及解決方法等）：見實(shí)驗(yàn)步驟中。1、由于心理、技術(shù)以及制度等原因，京津變量之間的影響往往具有滯帶效應(yīng)，滯帶變量模型在經(jīng)濟(jì)分析中具有重要作用。分布滯后模型和自回歸模型是兩種常見的滯后變量模型。2、分布滯后模型不能直接運(yùn)用OLS方法進(jìn)行估計(jì)，原因在于自由度損失、多重共線性和滯后長(zhǎng)度難于確定