1、滯后變量模型的應(yīng)用為了考察時間滯后影響，現(xiàn)給出2003年到2015年四川省房地產(chǎn)業(yè)固定資產(chǎn)投資與銷售額的統(tǒng)計數(shù)據(jù)1如下表所示:年份房地產(chǎn)業(yè)全社會固定資產(chǎn)投資Y （億元）商品房銷售額X （億元）2003639.79349.372004641.32330.482005839.78661.9520061089.81931.12200715731398.5920081791.021105.4620092749.422094.0520103157.52647.3420113993.873218.0420125017.533517.7220136479.114020.2720147546.913997.3

2、720157992.774199.841、設(shè)定模型，做局部調(diào)整假設(shè):（0< 1）則原模型變換為1來源于國家統(tǒng)計局網(wǎng)站地區(qū)數(shù)據(jù)利用Eviews軟件對上述模型進行估計，結(jié)果如下:|=| tquatiofi: UNtlTLEO WoFtrfiUe: 05:Untitied= X/if.Srcr Oh|?<t| Prin Na«Tp Freejp F農(nóng)RirtcaMSi ats Re «dsDpenckini vafi VSquaresDate- OtW-MS Time' W'24Sample (adjusted). 2004 2015he lulled

3、 obsgrvKins 12 d即 jul iTientsVariableCoeffciefit Std. Error i-Stabsdc Prob.0X 刃"75 769J50MS1290 79BE245653-0.51000(；U.13W16J.%43640 0044599 4557BE0.6223OQQ33ODDDCRujredAd|血ted R -squaredS F ofFTT怎Sum squared resdLog likelihood!FPrW(F-WlJStJC)0.9«2&94243 95253M86.381 2521fl637 B3730-000

4、000lb 屯an defjcideL wf S D cfependerrt var AXaike irfa crirericn Schwarz cntefcm Hannan Qminri entef Durbin Wrsfin 冷館t3S7ie?0 2633 685 14 04203 M t«32®139D715(3.954)(9.456)F=637.837DW=1.846從回歸結(jié)果可以求得原模型的回歸系數(shù)：，,，所以理想的固定投資函數(shù)為：2、設(shè)定模型，做自適應(yīng)預(yù)期假定：其中，0< <1。于是，原模型變換為：其中，。由于該模型存在解釋變量與滯后被解釋變量同期相

5、關(guān)的問題，不能直接進行 OLS估計，故采用工具變量法進行估計。回 Hurtion: UNTRIED 恥rWU豊 05:5憫邸曰 x甘柜丹2°4©切電白 Piirrt Ndm電FEt!生|亂hmak：冷住匸應(yīng) Stab |Dependent Variable Yh'etnoc Twa-Stace Les： SquaresDate. OI.Mj'lfi nme 19 5D&4呷尾仙悴檢f 2004 2015Ircluded cbservarticris 12 after adjustTwitshstumert spec ificatBr: C X Y&l

6、t;-1)丫滬剎6Codficeif Std Ernar t-StabstK ProbC-75 76525140 5651-0 51D0D606223X0S4BI33o冊4 把400033霞0 f9W240 QaMffl 9 455?w00000R-squared0 992994Mean dependent ar35 72 670Adpsted R-squaneC0 09138S D dependHni /ar2G33586$已 of悶essian243.6952Sum squared read9344% 3r-Eiad&tlc637.B373D jrbn-Watson s:ai1.G

7、J<i006ProbfRstat 妙coooooraSecond-Stage SSR53+4 &6 34sMisteOIKXKKKIInstrurnent 陽nk3即估計結(jié)果如下:(3.954)(9.456)F=637.8373、假設(shè)房地產(chǎn)業(yè)固定資產(chǎn)投資受當期和前三期銷售影響顯著，利用阿爾蒙多項式變換 估計以下有限分布滯后模型：a+利用阿爾蒙變換，令，原模型變換為a+利用Eviews軟件估計結(jié)果如下:r=l Equqtiorr EQ01 WbrMile: UbmTlfD:Llmiticd.S JC.卜 i.jjPrfi 札 dtoytdi jpiMTNmE*|R4tl»

8、 :EdiiW虻伽旳卅卓 加血2CRi_. <jiiIjjt o 3 nift.Siiplc2006 2015jAj'iciuded ctsefMUrK,IQiHtr 叫 3J7KHISVanaztieC-mHtc lentStd. EirarLSLrlisticProb.C164 57B17181F290 7013200 4753PDL01 2E363801S72821 407663D20B9PtiLoe 1Q1O6?+3jgciyejO4D1SPCLMoricaeo01825000 804M30 647BR-squared 994B7Dr.'ean dependert

9、l rar41391194Adjusled R-squared0 99200SS D deipeTdenr var25171323 F of regression226 0038Adike ihfti tilteiion13B6S82Sum 岡uored暗嘀d切冊特2Schwarz nrterian14 DBO65Log likelihood-5 7 9010HanFian-Cuinn zrifer13 B2705F-EtartsCic37S 2512Durbm -Watson stafr3DB1215Frot(F.srati5iic)0 000000Lag Dislribufcn of Xi

10、CoefViEStd Errorl-5<ati：>ric1 I0a 19743D22K408W7G1103Q1畸1 4075&20 5&QMD177683 0SBS31|31058130 252564 16663Sijpi of Lags2 09&78O0&7«4308Cna7164.5781 +(1.408 )( 0.902 )(0.605 )F=373.251DW=3.0814、設(shè)定模型In，做局部調(diào)整一一自適應(yīng)預(yù)期綜合模型進行估計:=Equadort EQOS WorkHlei U MTlTlEDirUntrtl ed- m i|叫

11、pi qc 衛(wèi)屮u | P rint ISd in*. Fi © 點.仙 zL* Fg g*t. 08 FtkJe Dependefii Va nabieMethcd. Least SquaresDale OtlG Trne 21.41Same(adpusted) 2006 2016Inckjripfi mbrvalKins 11 fjc HrrcntVariableCcfficierrtotri Ejtc”t-StdtisucProb.C0138926O20OH20634061060&3LQG(K)0 3719S?01064 4?34947D30Q101LO<3(Y( 1)041324J0.183317220153LW(Y(-2).2335700169490L370075o.2i oeR-squared0995675ctepenctent rar4IO0W18Adjusted R squared09933丹S D. <ftpend*nt /arG76O3BTSE. of rtgrciiscnD.OE2CM5Aka w rft crteii on2 446620Skit -iqiiared rsid臨卒商Schwar? cdterion”？伽伽Log hkeliihnDd17-15641H可nran-Quim izhIet2 537828F