用EVIEWS軟件對一組數(shù)據(jù)作協(xié)整分析

1、資產(chǎn)定價(jià)理論作業(yè)一:找一組你熟悉的經(jīng)濟(jì)數(shù)據(jù)，用EVIEWS軟件作協(xié)整分析。對2006年1月至2007年10月的中美貿(mào)易和中英貿(mào)易數(shù)據(jù)做協(xié)整分析如下:一，對變量CA和CE進(jìn)行單位根檢驗(yàn)令CA為中美貿(mào)易，CE為中英貿(mào)易，趨勢圖如下圖：CE數(shù)據(jù)如下圖:對原序列進(jìn)行ADF檢驗(yàn)，檢驗(yàn)結(jié)果如下圖:rni's-Series: CE V/orfcfilcf 幣英黃勇=Uri廿廿W.| Rroc I C»fc>jec±| PruFJertie：、| Namr | Fyew | 口 tz =Gczn；Eh-r 11 Cj曰口h | Wtts | EcjEnt |Aiiamente

2、d Dickey-Fuller Unil Root Test on CENull Hypothe sis: CE has a unat rootExog&Aous: Constant, Liine-ar TrendLag Length: 0 fAutomatic based on SIC. FJA.xlaG-4)t-StatlsticF母£).“Augmented Dic*-e Fli 11 er test statistic-1.8855020-6261Test critical values:1%« lvE5% level 10% level-4.467895-

3、3.644963-3.261452mgcKinn門 1 995. one-sided p-values.Augmented DlcFey-Fullsr Test EquationDspnderitVrishle： D(CE)Method: Least S口uaresDate: 05/13. 11 Time: 12:52sample i.adjusted 20O6UO2 sootmioIncluded cbser.tiQins 21 after gidijuiistments由上圖看出，CE為非平穩(wěn)序歹0,對其進(jìn)行一階差分檢驗(yàn)，檢驗(yàn)結(jié)果如下圖:由P值可以看出一階差分單位根檢驗(yàn)為平穩(wěn)序歹0,即 同

4、樣，對中美貿(mào)易進(jìn)行檢驗(yàn)，得到趨勢圖如下圖：CAI (1),是一階單整16,000,00014,000,00012,000,00010,000,0008,000,0006,000,0004,000,0002,000,000CA數(shù)據(jù)如下圖所示:F SeriK： CA Workfile:中美魂:UntitledVV lew j Pioc | Object | Pro口噌r七亡步Fri舟t r占E之：戶t。亡上尺 Default | %t EdiR-i-J-5mpl +/- ；j L-at6Last updated: OE/13/11 12:37Jfc.I2006T.1011007600.2006M0

5、21767241.2006U0328599922006L10439321122006M055027-1182006M066237619.2006TJ07742071420J5EiaSB0O912T2006L1091G2205S32006IM1011S239312006M11130S13232006U12144263452007NO11295421.2007L1022519228.2007M03368530.2007LI 0447216782007M0559651E1.2QO7MO6739149S.2007L1070030125.2007M081033000220071.109118622292

6、007M10先進(jìn)行ADF檢驗(yàn)，得到結(jié)果為:由P值判斷所得原序列不是平穩(wěn)序列，故進(jìn)行一階差分檢驗(yàn)得到結(jié)果如下:由上圖的P值可以看出，CA的一階差分序歹0為平'穩(wěn)的，即 C冬I (1),為一階 單整。兩個對象均為一階單整故進(jìn)行協(xié)整分析。二，用最小二乘法進(jìn)行估計(jì)。用最小二乘法對回歸模型：CA=b0+b1CE+u進(jìn)行估 計(jì)，運(yùn)用EVIEWSt下面數(shù)據(jù)進(jìn)行分析：CE:Tbntitiedl°Vie-' | ProcObject PrintMameFreeze | jDefault鈕rtEcht+"Smpl+/-|TitiabgceCA|obsC£CAJ2-ju

7、eL1C'1120673.011007600?2006?.10221037S.D1767241.20D6M03344123.0285999220Q6T.1044715S2.03932112.2006M05=96947.05027413._2D06M06736277.06237019.2006T/07879790.Q742D714.20D6M091050257.B809127.辨。引.1。9123042.102203320G6T.1101427779.11623931200614m1509687.1308132320Q6M121765683.1442634532007t.10116686

8、9.01296421.2OQ7L1Q2359064.0251922520U7L103和了以.034&&530.2007T.104677364 047215TB；20071.105856219.059851512007ti10l51037600.7391438.20Q7M071249231.3S3012520a7bl081469436.1033000220D7M091712116.1186222QI20071.1101955240.13335587Ii| 4得到結(jié)果為:ViewProc | Object | Print | Name | Freeze | Es ;i nia te

9、 ForecastStats | ResidsDependent variable: CA Method: Least Squares Date: 05/13；11 Time: 15:24 ampie： 2006U01 2007U10 Inctuded obsenations: 22.ariabjeCoeffi ci entStd. Errort-StatislJCProto.cC£233402 97.390830301306.00.2732780.7743726 4640?0.44760.0000R-squaredAdjusted R-squsredS.E. of resressi

10、onSum squared rssidLog likelihood F-statfsticProb(F-statisttc)0.972236 0 970847 713371 11.03E*13 -32J3 0325700.345? 0.000000Mean dependent ,ar S O dependentvar .kaie info criterion Schwarz criterion Hannan-Quinn criter. Durbin-Z/atson stat7100207.4207360.29 S938729 9930529 91723 0 208363二，檢驗(yàn)殘差序列的平穩(wěn)性

11、。 殘差趨勢圖為:1,500,000RESID1,000,000500,000-500,000-1,000,000-1,500,000殘差序列為:進(jìn)行ADF檢驗(yàn)，得到結(jié)果為:由P值看出結(jié)果是不平穩(wěn)的，故進(jìn)行一階單位根檢驗(yàn)，得到結(jié)果為:| Prou | ObiE匚t ProgrtiMS | Pirint| Nbee | EeeTie j Sample j QiEfTr EhMit| Graiph | Statskdent|Wui口nristifedl Dickey-FuiTeF umt RooFlst on D(U)NliII H/ipothesis D(U> has a Limit ro

12、otExogenous. Constan!Lag Length: 0 i Autom：atlc based on SIC. hl.AXLAG-4：t-Stati sticProb.x二.l曰mwntEe! OiuHm-FliMe teml: mtmtiwiiu-3.了了1 182O.D1D日Test critical valuer:1% Isvel-3.8085465% level-3.02053&10% level-2.650413*L1a：cKmncn i 199 6; (n岸一與id*。p-vluesAugmented Dicke/-Full er Te st EquationD

13、epe ndent Vanable: D(U.2；Method: Least SquaresDte: 05/13V11 Time： 15：27Sample (adjusted:), 2D06M03 20=a7M1OHncludGd observations. 20 sTter adjustments再進(jìn)行二階單位根檢驗(yàn)結(jié)果為:-T-|I,1 'J |"l.Series; U WorknFer CA;:Unbaed1J¥吧牌Prou | Object | Propef 為史占戶Tint: | Nam. | Free王亡Sample： | Gem Shee：U I 占也

14、仁! IdenkAngEerrteHd Di.lceyluHeT Unit Ftoot Tgt <ni ；0<11衛(wèi))r J mll Hypothesis : D(uP2)has a unit rootExogenous; Mone Lag Length: 0 (Automaticbased on SIC. hl.AXLAG=4)t-SlattsttcProfit.uainentea Dicke.-Fuller test statistic-7.2004720 0000Test critical values:1% level-2.6923585% level-1.96017110% level-1.807051"f.lacKinnon (1996 one-sided p-values./. aming- Probabilities and critical values calculated for 20 observat