论文首页哲学论文经济论文法学论文教育论文文学论文历史论文理学论文工学论文医学论文管理论文艺术论文 |
通货膨胀的影响因素分析
文主要通过对通货膨胀进行多因素分析,建立以度量通货膨胀的商品零售物价指数为应变量,以其它可量化影响因素为自变量的多元线性回归模型,对影响通货膨胀的各主要经济因素进行考察,说明它们对通货膨胀的影响程度,从而为我国避免严重的通货膨胀以确保经济的持续稳定发展提供理论依据。
关键词:通货膨胀 多因素分析 模型 计量经济学 检验 修正
一.问题提出
在中国经济高速增长的同时,一系列现象也值得我们深思,中国是否经济已经过热?投资增速过快,2003年全社会固定资产投资70073亿元,比上年增长26.1%,钢铁、电解铝、水泥行业过度投资愈演愈烈。货币信贷增长偏快,央行在2003年货币政策执行报告中曾预测2004年M2和M1分别增长17%左右,但是2004年2月末广义货币M2实际增长19.4% 狭义货币M1实际增长19.8%,明显超过了央行的预期。物价总水平继续上升,2004年居民消费价格指数,工业品出厂价格指数,原材料、燃料、动力购进价格指数较2003年都有3个指数点以上的增长。这些都说明我国面临着较大的通胀压力。通货膨胀的结果是严重的,不仅会扭曲商品市场的价格,使资源配置无效率,还同时会扭曲金融市场的价格,引起泡沫。
二.文献综述
1.货膨胀的意思是:物品和生产要素的价格普遍上升的时期。通货膨胀意味着一般价格水平的上涨。今天我们用价格指数即成千上万种产品的加权平均价格来计算通货膨胀。
(保罗·萨缪尔森《经济学》第十一版)
平均价格水平的上升,并不是任何一种特殊价格的上升。
(布拉德利·希勒《当代经济学》)
3.所谓通货膨胀率,实际上也就是物价指数(如居民消费物价指数)的年增长率。一般地,各国用以计算通货膨胀率的物价指数主要有消费物价指数(CPI)(或者零售物价指数RPI),批发物价指数(WPI)(或者生产者价格指数PPI),以及国内生产总值缩减指数(IPD)等三种。
t=(-0.431801) (-6.216710) (7.176346) (0.884635) (-4.986093) (13.77827)
=0.994450 DW=1.641975 F=609.1879
由以上结果可知,=0.994450,说明模型整体拟合得很好, 各因素对物价的解释程度高达99.445%;F=609.1879>F(18,4)=2.93 (显著性水平a=0.05),表明模型从整体上看物价指数与各解释变量间线形关系显著;
但是变量M(-1)参数的t值不显著,t=0.884635,而且G(-1)与F(-1)的参数值符号为负, 明显与经济意义不符, 根据变量显著性和方程显著性的综合判断, 可初步判断该模型存在多重共线性,需要进行修正。 内容来自www.nseac.com
七. 各种检验和修正
1.多重共线性检验和修正
(1)检验
计算各解释变量之间的简单相关系数,得相关系数矩阵:
G(-1) I(-1) M(-1) F(-1) Y(-1)
G(-1) 1
I(-1) 0.988020020169 1
M(-1) 0.93469553168 0.953971424757 1
F(-1) 0.937637199233 0.974600643179 0.943210798295 1
Y(-1) 0.911435296958 0.849934843755 0.766406750361 0.73233274266 1
由上表可见,各个解释变量间都存在高度相关性, 由模型回归结果也可看出, 尽管模型整体拟合较好,但M(-1)的参数t值不显著,G(-1)与F(-1)的参数符号与经济意义相悖, 表明模型确实存在严重的多重共线性,需要进行修正
(2)修正
用逐步回归法:
ⅰ.运用OLS方法逐一求Y对各个解释变量的回归, 结合经济意义和统计检验选出拟合效果最好的一元线形回归方程:
经分析, 在五个一元回归模型的中Y与Y(-1)的线形关系最强,拟合程度最好,因此,纳入Y(-1)得模型①:
①
ⅱ.逐步回归,
将其余解释变量一一代入①式得如下四个模型:
根据回归的结果, 对比分析得: 纳入G(-1)后使得提高的最多,且符合经济意义, Y(-1)显著, 自身的显著性相对于其它变量也要强一些, 所以在模型①中再纳入G(-1)得模型②
②
将其余解释变量再一一代入②式得如下三个模型:
根据回归的结果, 对比分析得:纳入I(-1)后使得提高的最多,且符合经济意义,t值检验也都是显著的, 所以在模型中再纳入I(-1)得模型③
③
将剩余解释变量再一一代入③式得如下两个模型:
将M(-1)纳入后,虽然有一点点的提高, 但是对其它参数的符号和数值没有什么影响, 而且M(-1)的t值还显著,所以可以舍去M(-1)
将F(-1)纳入后, 虽然有一些些的提高, 但是F(-1)参数的符号为负, 不符合经济意义, 所以还是舍去F(-1) 大学排名
综上可得, Y对G(-1), I(-1), Y(-1)的回归模型为最优, 即通过多重共线性的修正得到的最优模型为模型③
2.异方差的检验与修正
(1)检验
a.White检验
有交叉项的White检验,结果如下:
White Heteroskedasticity Test:
F-statistic 2.868072 Probability 0.041593
Obs*R-squared 10.29633 Probability 0.083111
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 06/05/05 Time: 22:29
Sample: 1982 2004
Included observations: 23
Variable Coefficient Std. Error t-Statistic Prob.
C -1609.676 794.0239 -2.027239 0.0637
Y(-1) 24.29959 13.29177 1.828168 0.0906
Y(-1)^2 -0.045494 0.049106 -0.926439 0.3711
Y(-1)*I(-1) 0.000256 0.001779 0.144030 0.8877
Y(-1)*G(-1) 3.86E-05 0.000921 0.041888 0.9672
I(-1) 0.303240 0.275276 1.101585 0.2906
I(-1)^2 9.88E-06 1.34E-05 0.735703 0.4750
I(-1)*G(-1) -1.18E-05 1.52E-05 -0.773686 0.4530
G(-1) -0.184846 0.133734 -1.382197 0.1902
G(-1)^2 3.14E-06 4.31E-06 0.727475 0.4798
R-squared 0.665058 Mean dependent var 140.6020
Adjusted R-squared 0.433175 S.D. dependent var 157.0554
S.E. of regression 118.2436 Akaike info criterion 12.68239
Sum squared resid 181760.2 Schwarz criterion 13.17608
Log likelihood -135.8475 F-statistic 2.868072
Durbin-Watson stat 1.933704 Prob(F-statistic) 0.041593
(科教论文网 Lw.nsEAc.com编辑整理)
没有交叉项的White检验,结果如下:
White Heteroskedasticity Test:
F-statistic 3.287564 Probability 0.026532
Obs*R-squared 6.69920 Probability 0.048069
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 06/05/05 Time: 22:30
Sample: 1982 2004
Included observations: 23
Variable Coefficient Std. Error t-Statistic Prob.
C -706.6808 469.3675 -1.505602 0.1517
Y(-1) 9.936950 6.508144 1.526849 0.1463
Y(-1)^2 -0.012226 0.006322 -1.933787 0.0710
I(-1) 0.182330 0.076852 2.372483 0.0305
I(-1)^2 -1.65E-06 9.46E-07 -1.746969 0.0998
G(-1) -0.087019 0.058422 -1.489487 0.1558
G(-1)^2 2.92E-07 3.35E-07 0.871253 0.3965
R-squared 0.552139 Mean dependent var 140.6020
Adjusted R-squared 0.384191 S.D. dependent var 157.0554
S.E. of regression 123.2469 Akaike info criterion 12.71205
Sum squared resid 243036.6 Schwarz criterion 13.05763
Log likelihood -139.1885 F-statistic 3.287564
Durbin-Watson stat 2.028429 Prob(F-statistic) 0.026532
由上可得, Obs*R-squared=1.412724<(6)=7.81473, 所以接受H0,拒绝H1,表明此模型随机误差U不存在异方差。
b.ARCH检验
直接用Eviews检验,结果如下:
ARCH Test:
F-statistic 1.210277 Probability 0.337999
(科教范文网http://fw.NSEAC.com编辑发布)
Log likelihood -127.9953 F-statistic 1.210277
Durbin-Watson stat 2.062402 Prob(F-statistic) 0.337999
由上可得, Obs*R-squared=3.699110< (3)=7.81473,所以接受H0,拒绝H1, 表明此模型随机误差U不存在异方差。
由上述两种检验方法都表明该模型确实不存在异方差.
3.自相关的检验与修正
(1)检验
由于模型中有应变量Y的滞后期Y(-1),故不能用DW检验,应该用德宾h—检验
由多重共线性修正后的模型结果得: DW=0.967524,Var()=0.107154^2,n=23,
则=2.886> h(a/2)=1.96 (a=0.05)
因此拒绝原假设=0,说明模型存在正的一阶自相关。
(2)修正
先利用对数线形回归修正自相关,得模型④:
④
DW=0.972403,Var()=0.146176^2,n=23,
则=3.455> h(a/2)=1.96 (a=0.05) 中国大学排名
因此拒绝原假设=0,说明对数模型仍然存在正的一阶自相关
同时考虑Cochrane-Orcutt迭代法,得模型⑤:
⑤
DW=1.615575 , Var()=0.374280^2,n=22
则=0.625<h(a/2)=1.96 (a=0.05)
因此接受原假设=0,说明修正后的模型不存一阶自相关了
再对模型⑤进行ARCH检验:
ARCH Test:
F-statistic 0.258752 Probability 0.853939
Obs*R-squared 0.934879 Probability 0.817004
Test Equation:
Dependent Variable: RESID^2
Method: Least Squares
Date: 06/05/05 Time: 23:47
Sample(adjusted): 1986 2004
Included observations: 19 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 0.000862 0.000756 1.140467 0.2720
RESID^2(-1) -0.043121 0.256355 -0.168210 0.8687
RESID^2(-2) 0.188130 0.253208 0.742987 0.4690
RESID^2(-3) 0.128607 0.250396 0.513613 0.6150
R-squared 0.049204 Mean dependent var 0.001225
Adjusted R-squared -0.140955 S.D. dependent var 0.002063
S.E. of regression 0.002204 Akaike info criterion -9.212547
Sum squared resid 7.29E-05 Schwarz criterion -9.013718
Log likelihood 91.51920 F-statistic 0.258752
Durbin-Watson stat 1.912210 Prob(F-statistic) 0.853939
由上可得, Obs*R-squared=0.934879< (3)=7.81473, 所以接受H0, 拒绝H1, 表明此模型随机误差U不存在异方差.
综上, 自相关修正后的模型⑤中、F值皆有所提高,t值除常数项外都显著, 而且不存在异方差和自相关,故模型⑤为最佳
八.最终模型和结论
模型表明, =0.992877,模型拟合很好, 即当期物价变动有99.29%可由滞后一期的GDP, 滞后一期的投资和滞后一期的物价共同解释,且滞后一期GDP, 滞后一期投资与滞后一期物价对当期物价的分别影响都显著,t值分别为2.41218 , 2.457413, 2.50480
滞后一期GDP每变动1%,引起当期物价变动0.3619%,可见通货膨胀是经济发展中不可避免的现象,所以我们对通胀的态度应该是在保证经济发展的前提下积极治理,而不应该过于敏感以至于造成经济减缓甚至停滞;滞后一期投资变动1%,引起当期物价变动0.333%说明应该更加重视投资因素对通货膨胀的作用,更加重视通过控制投资来控制通涨;滞后一期物价指数变动1%,引起当期物价变动0.842%说明通胀的治理是一个长期的过程,不可能通过某一段时间的治理就能得到明显的改善.
2.结论
从该模型可以看出, 模型的可决系数较高, 可见我们对于解释变量的选择是比较全面的, 它们对应变量的联合影响程度很大, 虽然修正后解释变量显著性有所下降, 但依然能符合检验要求, 可知通货膨胀是多种经济因素共同作用的结果,单独一个影响因素的作用也许不是那么明显。
还需要说明的是, 在模型建立之初,考虑到我国还处在市场经济初级,经济发展不太成熟,政府干预较多,政策作用较为明显,我们曾尝试将政府政策和政府行为作为变量纳入模型中,但由于其自身的复杂性,而且很难量化, 显然是不可行的,所以最终舍弃。但从模型的最后结果可以看出,可决系数是很高的, 达到0.992877, 说明没有纳入政府干预因素,做出来的模型效果也不错,这表明政府的干预作用并不像我们想象的那么大, 可以侧面反映我国市场经济发展的迅速和健康程度, 现在有很多国家对我国市场经济体制持怀疑态度, 该模型也为对这种怀疑的反驳提供了一定的理论依据。
(科教论文网 lw.nseaC.Com编辑发布)
参考文献:
保罗·萨缪尔森《经济学》第十一版
布拉德利·希勒《当代经济学》
«通货膨胀问题研究»«中国物价»2005.01
«需求推动角度考虑通货膨胀成因的实证分析»
«我国通货膨胀的成因分析»«天津市职工现代企业管理学院学报»2004.12第四期
附表:
数据
Y G I M F
110.7 4860.3 961 2299.96 27.08
112.8 5301.8 1230.4 2676.94 69.86
114.5 5957.4 1430.1 3193.57 89.01
117.7 7206.7 1832.9 4442.88 82.2
128.1 8989.1 2543.2 5198.9 26.44
135.8 10201.4 3210.6 6720.9 20.72
145.7 11954.5 3791.7 8330.9 29.23
172.7 14922.3 4753.8 100099.6 33.72
203.4 16917.8 4410.4 11949.6 55.5
207.7 18598.4 4517 15293.4 110.93
213.7 21662.5 5594.5 19349.9 217.12
225.2 26651.9 8080.1 25402.2 194.43
254.9 34560.5 13072.3 34879.8 211.99 内容来自www.nseac.com
310.2 46670 17042.1 46923.5 516.2
356.1 57494.9 20019.3 60750.5 735.97
377.8 66850.5 22913.5 76094.9 1050.29
380.8 73142.7 24941.1 90995.3 1398.9
370.9 76967.2 28406.2 104498.5 1449.6
359.8 80579.4 29854.7 119897.9 1546.75
354.4 88254 32917.7 134610.4 1655.74
351.6 95727.9 37213.5 158301.9 2121.65
347 103935.3 43499.91 185007 2864.07
346.7 116603.2 55566.61 221222.8 4032.51
356.4 136584.3 70073 253000 6099
Eview回归结果
Variable Coefficient Std. Error t-Statistic Prob.
C 144.7871 14.05072 10.30460 0.0000
PDL01 0.008356 0.002216 3.770521 0.0015
PDL02 0.003669 0.002087 1.758535 0.0966
PDL03 -0.006574 0.002127 -3.090887 0.0066
R-squared 0.882584 Mean dependent var 272.2190
Adjusted R-squared 0.861864 S.D. dependent var 95.94448
S.E. of regression 35.65934 Akaike info criterion 10.15554
Sum squared resid 21617.00 Schwarz criterion 10.35450
Log likelihood -102.6332 F-statistic 42.59499
Durbin-Watson stat 0.455755 Prob(F-statistic) 0.000000
Lag Distribution of G i Coefficient Std. Error T-Statistic
* . | 0 -0.00189 0.00212 -0.89143
. *| 1 0.00836 0.00222 3.77052
. * | 2 0.00545 0.00209 2.60496 内容来自www.nseac.com
* . | 3 -0.01060 0.00257 -4.12947
Sum of Lags 0.00132 0.00035 3.81522
Variable Coefficient Std. Error t-Statistic Prob.
C 175.2020 16.27290 10.76649 0.0000
PDL01 0.023168 0.006445 3.594945 0.0022
PDL02 0.018087 0.005192 3.483436 0.0028
PDL03 -0.020560 0.005944 -3.459201 0.0030
R-squared 0.787022 Mean dependent var 272.2190
Adjusted R-squared 0.749438 S.D. dependent var 95.94448
S.E. of regression 48.02617 Akaike info criterion 10.75101
Sum squared resid 39210.72 Schwarz criterion 10.94997
Log likelihood -108.8856 F-statistic 20.94014
Durbin-Watson stat 0.519259 Prob(F-statistic) 0.000006
Lag Distribution of I i Coefficient Std. Error T-Statistic
* . | 0 -0.01548 0.00494 -3.13514
. *| 1 0.02317 0.00644 3.59495
. * | 2 0.02070 0.00553 3.74316
* . | 3 -0.02290 0.00798 -2.86842
Sum of Lags 0.00549 0.00145 3.79400
Variable Coefficient Std. Error t-Statistic Prob.
C 212.7083 21.06730 10.09661 0.0000
PDL01 0.157300 0.099350 1.583297 0.1318
(转载自http://zw.NSEaC.com科教作文网)
Variable Coefficient Std. Error t-Statistic Prob.
C 18.48038 9.019337 2.048973 0.0572
PDL01 -0.323368 0.139850 -2.312256 0.0344
PDL02 -1.108375 0.148189 -7.479468 0.0000
PDL03 0.745015 0.141633 5.260197 0.0001
R-squared 0.983782 Mean dependent var 279.9450
Adjusted R-squared 0.980741 S.D. dependent var 91.48969
S.E. of regression 12.69667 Akaike info criterion 8.097413
Variable Coefficient Std. Error t-Statistic Prob.
C 197.2142 23.24994 8.482353 0.0000
PDL01 0.000419 0.000420 0.998934 0.3318
PDL02 -8.60E-05 0.000522 -0.164672 0.8711
PDL03 -7.64E-05 0.000395 -0.193215 0.8491
R-squared 0.527836 Mean dependent var 272.2190
Adjusted R-squared 0.444512 S.D. dependent var 95.94448
S.E. of regression 71.50841 Akaike info criterion 11.54715
Sum squared resid 86928.70 Schwarz criterion 11.74611
Log likelihood -117.2451 F-statistic 6.334803
Durbin-Watson stat 0.126593 Prob(F-statistic) 0.004425
Lag Distribution of M i Coefficient Std. Error T-Statistic
. *| 0 0.00043 0.00060 0.71553
(科教作文网http://zw.ΝsΕAc.com发布)
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 12:15
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C -4.020024 9.309908 -0.431801 0.6713
G(-1) -0.004353 0.000700 -6.216710 0.0000
I(-1) 0.012670 0.001766 7.176346 0.0000
M(-1) 8.80E-05 9.95E-05 0.884635 0.3887
F(-1) -0.066149 0.013267 -4.986093 0.0001
Y(-1) 1.193877 0.086649 13.77827 0.0000
R-squared 0.994450 Mean dependent var 258.4304
Adjusted R-squared 0.992817 S.D. dependent var 102.2527
S.E. of regression 8.665968 Akaike info criterion 7.376142
Sum squared resid 1276.683 Schwarz criterion 7.672358
Log likelihood -78.82564 F-statistic 609.1879
Durbin-Watson stat 1.641975 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:02
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
(科教作文网http://zw.NSEaC.com编辑发布)
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:06
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 173.4674 18.01594 9.628553 0.0000
I(-1) 0.005313 0.000811 6.552266 0.0000
R-squared 0.671527 Mean dependent var 258.4304
Adjusted R-squared 0.655885 S.D. dependent var 102.2527
S.E. of regression 59.98277 Akaike info criterion 11.10893
Sum squared resid 75556.58 Schwarz criterion 11.20767
Log likelihood -125.7527 F-statistic 42.93219
Durbin-Watson stat 0.132891 Prob(F-statistic) 0.000002
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:07
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 186.6627 20.64986 9.039414 0.0000
M(-1) 0.001148 0.000231 4.977152 0.0001
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:10
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 205.6000 19.99478 10.28269 0.0000
F(-1) 0.065540 0.015117 4.335503 0.0003
R-squared 0.472317 Mean dependent var 258.4304
Adjusted R-squared 0.447189 S.D. dependent var 102.2527
S.E. of regression 76.02622 Akaike info criterion 11.58297
Sum squared resid 121379.7 Schwarz criterion 11.68171
Log likelihood -131.2042 F-statistic 18.79658
Durbin-Watson stat 0.136589 Prob(F-statistic) 0.000291
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:10
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 18.96334 9.366769 2.024534 0.0558
Y(-1) 0.966576 0.034957 27.65036 0.0000
R-squared 0.973267 Mean dependent var 258.4304
Adjusted R-squared 0.971994 S.D. dependent var 102.2527 (科教论文网 lw.nSeAc.com编辑发布)
S.E. of regression 17.11203 Akaike info criterion 8.600382
Sum squared resid 6149.252 Schwarz criterion 8.699120
Log likelihood -96.90439 F-statistic 764.5423
Durbin-Watson stat 0.543100 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:13
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 5.579119 12.40127 0.449883 0.6576
G(-1) -0.000367 0.000233 -1.578704 0.1301
Y(-1) 1.084702 0.082095 13.21273 0.0000
R-squared 0.976229 Mean dependent var 258.4304
Adjusted R-squared 0.973852 S.D. dependent var 102.2527
S.E. of regression 16.53462 Akaike info criterion 8.569897
Sum squared resid 5467.872 Schwarz criterion 8.718005
Log likelihood -95.55382 F-statistic 410.6823
Durbin-Watson stat 0.676758 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:15
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 11.92845 11.65612 1.023364 0.3184
I(-1) -0.000444 0.000439 -1.012901 0.3232
Y(-1) 1.023658 0.066305 15.43854 0.0000
R-squared 0.974571 Mean dependent var 258.4304
Adjusted R-squared 0.972028 S.D. dependent var 102.2527
(科教作文网http://zw.nseAc.com)
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:17
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 14.58319 10.63195 1.371638 0.1854
M(-1) -7.72E-05 8.71E-05 -0.886734 0.3858
Y(-1) 1.003749 0.054699 18.35036 0.0000
R-squared 0.974278 Mean dependent var 258.4304
Adjusted R-squared 0.971706 S.D. dependent var 102.2527
S.E. of regression 17.19977 Akaike info criterion 8.648776
Sum squared resid 5916.640 Schwarz criterion 8.796884
Log likelihood -96.46093 F-statistic 378.7742
Durbin-Watson stat 0.674830 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 13:20
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 11.29909 10.45499 1.080737 0.2927
F(-1) -0.007245 0.004857 -1.491514 0.1514
Y(-1) 1.021083 0.049902 20.46192 0.0000
R-squared 0.975943 Mean dependent var 258.4304
Adjusted R-squared 0.973537 S.D. dependent var 102.2527 本文来自中国科教评价网
S.E. of regression 16.63390 Akaike info criterion 8.581870
Sum squared resid 5533.731 Schwarz criterion 8.729978
Log likelihood -95.69150 F-statistic 405.6756
Durbin-Watson stat 0.649189 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 14:14
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C -20.71999 12.18349 -1.700662 0.1053
G(-1) 0.004066 0.001037 3.919966 0.0009
Y(-1) 1.393955 0.107154 13.00885 0.0000
I(-1) 0.006849 0.001890 3.622942 0.0018
R-squared 0.985941 Mean dependent var 258.4304
Adjusted R-squared 0.983721 S.D. dependent var 102.2527
S.E. of regression 13.04616 Akaike info criterion 8.131636
Sum squared resid 3233.845 Schwarz criterion 8.329113
Log likelihood -89.51381 F-statistic 444.1572
Durbin-Watson stat 0.967524 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 14:17
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C -1.270748 13.84150 -0.091807 0.9278
Y(-1) 1.149056 0.100705 11.41008 0.0000
G(-1) -0.000871 0.000516 -1.688798 0.1076
M(-1) 0.000203 0.000186 1.092974 0.2881
R-squared 0.977635 Mean dependent var 258.4304
(转载自http://zw.NSEaC.com科教作文网)
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 14:18
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C 4.624315 17.40470 0.265694 0.7933
Y(-1) 1.095854 0.162296 6.752200 0.0000
G(-1) -0.000437 0.000901 -0.485170 0.6331
F(-1) 0.001503 0.018698 0.080383 0.9368
R-squared 0.976237 Mean dependent var 258.4304
Adjusted R-squared 0.972485 S.D. dependent var 102.2527
S.E. of regression 16.96128 Akaike info criterion 8.656514
Sum squared resid 5466.013 Schwarz criterion 8.853991
Log likelihood -95.54991 F-statistic 260.1894
Durbin-Watson stat 0.681311 Prob(F-statistic) 0.000000
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 14:26
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C -23.41426 12.89907 -1.815189 0.0862
Y(-1) 1.416942 0.113165 12.52102 0.0000
G(-1) -0.004198 0.001067 -3.935619 0.0010
Dependent Variable: Y
Method: Least Squares
Date: 06/05/05 Time: 14:28
Sample(adjusted): 1982 2004
Included observations: 23 after adjusting endpoints
Variable Coefficient Std. Error t-Statistic Prob.
C -1.701348 8.879219 -0.191610 0.8502
G(-1) -0.004247 0.000686 -6.193523 0.0000
Y(-1) 1.173664 0.083076 14.12759 0.0000
I(-1) 0.012922 0.001732 7.461501 0.0000
F(-1) -0.066643 0.013175 -5.058417 0.0001
R-squared 0.994194 Mean dependent var 258.4304
Adjusted R-squared 0.992904 S.D. dependent var 102.2527
S.E. of regression 8.613471 Akaike info criterion 7.334192
Sum squared resid 1335.454 Schwarz criterion 7.581039
Log likelihood -79.34321 F-statistic 770.5972
Durbin-Watson stat 1.707065 Prob(F-statistic) 0.000000
Variable Coefficient Std. Error t-Statistic Prob.
C 1.463094 0.315764 4.633498 0.0002
LN_G(-1) 0.796564 0.191679 4.155717 0.0005 (科教论文网 Lw.nsEAc.com编辑整理)
LN_I(-1) 0.529338 0.124035 4.267665 0.0004
LN_Y(-1) 1.356134 0.146176 9.277402 0.0000
R-squared 0.990185 Mean dependent var 5.465324
Adjusted R-squared 0.988635 S.D. dependent var 0.451488
S.E. of regression 0.048132 Akaike info criterion -3.072961
Sum squared resid 0.044017 Schwarz criterion -2.875483
Log likelihood 39.33905 F-statistic 638.9099
Durbin-Watson stat 0.972403 Prob(F-statistic) 0.000000
Dependent Variable: LN_Y
Method: Least Squares
Date: 06/05/05 Time: 15:56
Sample(adjusted): 1983 2004
Included observations: 22 after adjusting endpoints
Convergence achieved after 13 iterations
Variable Coefficient Std. Error t-Statistic Prob.
C 1.584930 1.047378 1.513235 0.1486
LN_G(-1) 0.361869 0.251085 2.441218 0.1677
LN_I(-1) 0.332900 0.135468 2.457413 0.0250
LN_Y(-1) 0.842310 0.374280 2.250480 0.0379
AR(1) 0.813389 0.311174 2.613931 0.0181
R-squared 0.992877 Mean dependent var 5.498947
Adjusted R-squared 0.991201 S.D. dependent var 0.431635
S.E. of regression 0.040489 Akaike info criterion -3.378836
Sum squared resid 0.027870 Schwarz criterion -3.130871
Log likelihood 42.16719 F-statistic 592.3855
Durbin-Watson stat 1.615575 Prob(F-statistic) 0.000000
Inverted AR Roots .81