我国广义货币供应量分布滞后模型
货币主义学派认为，产生通货膨胀的必要条件是货币的超量供应。物价变动与货币供应量的变化有着较为密切的联系，但是二者之间的关系不是瞬时的，货币供应量的变化对物价的影响存在一定时滞。有研究表明，西方国家的通货膨胀时滞大约为2—3个季度。
在中国，大家普遍认同货币供给的变化对物价具有滞后影响，但滞后期究竟有多长，还存在不同的认识。下面采集1996－2005年全国广义货币供应量和物价指数的月度数据（见表6-1）对这一问题进行研究。
表6-1 1996－2005年全国广义货币供应量及物价指数月度数据
| 月度 |
广义货币M2 (千亿元） |
广义货币增长量M2z （千亿元） |
居民消费价格同比指数tbzs |
|
月度 |
广义货币M2 (千亿元） |
广义货币增长量M2z （千亿元） |
居民消费价格同比指数tbzs |
| Jan-96 |
58.401 |
|
|
|
Oct-00 |
129.522 |
-0.9518 |
100 |
| Feb-96 |
63.778 |
5.377 |
109.3 |
|
Nov-00 |
130.9941 |
1.4721 |
101.3 |
| Mar-96 |
64.511 |
0.733 |
109.8 |
|
Dec-00 |
134.6103 |
3.6162 |
101.5 |
| Apr-96 |
65.723 |
1.212 |
109.7 |
|
Jan-01 |
137.5436 |
2.9333 |
101.2 |
| May-96 |
66.88 |
1.157 |
108.9 |
|
Feb-01 |
136.2102 |
-1.3334 |
100 |
| Jun-96 |
68.132 |
1.252 |
108.6 |
|
Mar-01 |
138.7445 |
2.5343 |
100.8 |
| Jul-96 |
69.346 |
1.214 |
108.3 |
|
Apr-01 |
139.9499 |
1.2054 |
101.6 |
| Aug-96 |
72.309 |
2.963 |
108.1 |
|
May-01 |
139.0158 |
-0.9341 |
101.7 |
| Sep-96 |
69.643 |
-2.666 |
107.4 |
|
Jun-01 |
147.8097 |
8.7939 |
101.4 |
| Oct-96 |
73.1522 |
3.5092 |
107 |
|
Jul-01 |
149.2287 |
1.419 |
101.5 |
| Nov-96 |
74.142 |
0.9898 |
106.9 |
|
Aug-01 |
149.9418 |
0.7131 |
101 |
| Dec-96 |
76.0949 |
1.9529 |
107 |
|
Sep-01 |
151.8226 |
1.8808 |
99.9 |
| Jan-97 |
78.648 |
2.5531 |
105.9 |
|
Oct-01 |
151.4973 |
-0.3253 |
100.2 |
| Feb-97 |
78.998 |
0.35 |
105.6 |
|
Nov-01 |
154.0883 |
2.591 |
99.7 |
| Mar-97 |
79.889 |
0.891 |
104 |
|
Dec-01 |
158.3019 |
4.2136 |
99.7 |
| Apr-97 |
80.818 |
0.929 |
103.2 |
|
Jan-02 |
159.6393 |
1.3374 |
99 |
| May-97 |
81.151 |
0.333 |
102.8 |
|
Feb-02 |
160.9356 |
1.2963 |
100 |
| Jun-97 |
82.789 |
1.638 |
102.8 |
|
Mar-02 |
164.0646 |
3.129 |
99.2 |
| Jul-97 |
83.46 |
0.671 |
102.7 |
|
Apr-02 |
164.5706 |
0.506 |
98.7 |
| Aug-97 |
84.746 |
1.286 |
101.9 |
|
May-02 |
166.061 |
1.4904 |
98.9 |
| Sep-97 |
85.892 |
1.146 |
101.8 |
|
Jun-02 |
169.6012 |
3.5402 |
99.2 |
| Oct-97 |
86.644 |
0.752 |
101.5 |
|
Jul-02 |
170.8511 |
1.2499 |
99.1 |
| Nov-97 |
87.59 |
0.946 |
101.1 |
|
Aug-02 |
173.2509 |
2.3998 |
99.3 |
| Dec-97 |
90.9953 |
3.4053 |
100.4 |
|
Sep-02 |
176.9824 |
3.7315 |
99.3 |
| Jan-98 |
92.2114 |
1.2161 |
100.3 |
|
Oct-02 |
177.2942 |
0.3118 |
99.2 |
| Feb-98 |
92.024 |
-0.1874 |
99.9 |
|
Nov-02 |
179.7363 |
2.4421 |
99.3 |
| Mar-98 |
92.015 |
-0.009 |
100.7 |
|
Dec-02 |
185.0073 |
5.271 |
99.6 |
| Apr-98 |
92.662 |
0.647 |
99.7 |
|
Jan-03 |
190.4883 |
5.481 |
100.4 |
| May-98 |
93.936 |
1.274 |
99 |
|
Feb-03 |
190.1084 |
-0.3799 |
100.2 |
| Jun-98 |
94.658 |
0.722 |
98.7 |
|
Mar-03 |
194.4873 |
4.3789 |
100.9 |
| Jul-98 |
96.314 |
1.656 |
98.6 |
|
Apr-03 |
196.1301 |
1.6428 |
101 |
| Aug-98 |
97.299 |
0.985 |
98.6 |
|
May-03 |
199.5052 |
3.3751 |
100.7 |
| Sep-98 |
99.795 |
2.496 |
98.5 |
|
Jun-03 |
204.9314 |
5.4262 |
100.3 |
| Oct-98 |
100.8752 |
1.0802 |
98.9 |
|
Jul-03 |
206.1931 |
1.2617 |
100.5 |
| Nov-98 |
102.229 |
1.3538 |
98.8 |
|
Aug-03 |
210.5919 |
4.3988 |
100.9 |
| Dec-98 |
104.4985 |
2.2695 |
99 |
|
Sep-03 |
213.5671 |
2.9752 |
101.1 |
| Jan-99 |
105.5 |
1.0015 |
98.8 |
|
Oct-03 |
214.4694 |
0.9023 |
101.8 |
| Feb-99 |
107.778 |
2.278 |
98.7 |
|
Nov-03 |
216.3517 |
1.8823 |
103 |
| Mar-99 |
108.438 |
0.66 |
98.2 |
|
Dec-03 |
221.2228 |
4.8711 |
103.2 |
| Apr-99 |
109.218 |
0.78 |
97.8 |
|
Jan-04 |
225.10193 |
3.87913 |
103.2 |
| May-99 |
110.061 |
0.843 |
97.8 |
|
Feb-04 |
227.05072 |
1.94879 |
102.1 |
| Jun-99 |
111.363 |
1.302 |
97.9 |
|
Mar-04 |
231.6546 |
4.60388 |
103 |
| Jul-99 |
111.414 |
0.051 |
98.6 |
|
Apr-04 |
233.62786 |
1.97326 |
103.8 |
| Aug-99 |
112.827 |
1.413 |
98.7 |
|
May-04 |
234.8424 |
1.21454 |
104.4 |
| Sep-99 |
115.079 |
2.252 |
99.2 |
|
Jun-04 |
238.42749 |
3.58509 |
105 |
| Oct-99 |
115.39 |
0.311 |
99.4 |
|
Jul-04 |
234.8424 |
-3.58509 |
105.3 |
| Nov-99 |
116.559 |
1.169 |
99.1 |
|
Aug-04 |
239.72919 |
4.88679 |
105.3 |
| Dec-99 |
119.898 |
3.339 |
99 |
|
Sep-04 |
243.757 |
4.02781 |
105.2 |
| Jan-00 |
121.22 |
1.322 |
99.8 |
|
Oct-04 |
243.74 |
-0.017 |
104.3 |
| Feb-00 |
121.5834 |
0.3634 |
100.7 |
|
Nov-04 |
247.13558 |
3.39558 |
102.8 |
| Mar-00 |
122.5807 |
0.9973 |
99.8 |
|
Dec-04 |
253.2077 |
6.07212 |
102.4 |
| Apr-00 |
124.1219 |
1.5412 |
99.7 |
|
Jan-05 |
257.75283 |
4.54513 |
101.9 |
| May-00 |
124.0533 |
-0.0686 |
100.1 |
|
Feb-05 |
259.3561 |
1.60327 |
103.9 |
| Jun-00 |
126.6053 |
2.552 |
100.5 |
|
Mar-05 |
264.5889 |
5.2328 |
102.7 |
| Jul-00 |
126.3239 |
-0.2814 |
100.5 |
|
Apr-05 |
266.99266 |
2.40376 |
101.8 |
| Aug-00 |
127.79 |
1.4661 |
100.3 |
|
May-05 |
269.2294 |
2.23674 |
101.8 |
| Sep-00 |
130.4738 |
2.6838 |
100 |
|
|
|
|
|
数据来源：中国经济统计数据库，http://db.cei.gov.cn/。
为了考察货币供应量的变化对物价的影响，我们用广义货币M2的月增长量M2Z作为解释变量，以居民消费价格月度同比指数TBZS为被解释变量进行研究。首先估计如下回归模型
得如下回归结果（表7.5）。
表7.5
| Dependent Variable: TBZS |
| Method: Least Squares |
| Date: 07/03/05 Time: 17:10 |
| Sample(adjusted): 1996:02 2005:05 |
| Included observations: 112 after adjusting endpoints |
| Variable |
Coefficient |
Std. Error |
t-Statistic |
Prob. |
| C |
101.4356 |
0.397419 |
255.2358 |
0.0000 |
| M2Z |
0.068371 |
0.151872 |
0.450190 |
0.6535 |
| R-squared |
0.001839 |
Mean dependent var |
101.5643 |
| Adjusted R-squared |
-0.007235 |
S.D. dependent var |
2.911111 |
| S.E. of regression |
2.921623 |
Akaike info criterion |
4.999852 |
| Sum squared resid |
938.9472 |
Schwarz criterion |
5.048396 |
| Log likelihood |
-277.9917 |
F-statistic |
0.202671 |
| Durbin-Watson stat |
0.047702 |
Prob(F-statistic) |
0.653460 |
从回归结果来看，M2Z的t统计量值不显著，表明当期货币供应量的变化对当期物价水平的影响在统计意义上不明显。为了分析货币供应量变化影响物价的滞后性，我们做滞后6个月的分布滞后模型的估计，在Eviews工作文档的方程设定窗口中，输入
TBZS C M2Z M2Z(-1) M2Z(-2) M2Z(-3) M2Z(-4) M2Z(-5) M2Z(-6)
结果见表7.6。
表7.6
| Dependent Variable: TBZS |
| Method: Least Squares |
| Date: 07/03/05 Time: 17:09 |
| Sample(adjusted): 1996:08 2005:05 |
| Included observations: 106 after adjusting endpoints |
| Variable |
Coefficient |
Std. Error |
t-Statistic |
Prob. |
| C |
100.0492 |
0.584318 |
171.2240 |
0.0000 |
| M2Z |
-0.011037 |
0.140613 |
-0.078493 |
0.9376 |
| M2Z(-1) |
0.016169 |
0.137998 |
0.117166 |
0.9070 |
| M2Z(-2) |
0.053044 |
0.136808 |
0.387723 |
0.6991 |
| M2Z(-3) |
0.028679 |
0.143155 |
0.200333 |
0.8416 |
| M2Z(-4) |
0.130825 |
0.139183 |
0.939951 |
0.3496 |
| M2Z(-5) |
0.137794 |
0.142502 |
0.966965 |
0.3359 |
| M2Z(-6) |
0.248778 |
0.143394 |
1.734924 |
0.0859 |
| R-squared |
0.055557 |
Mean dependent var |
101.1377 |
| Adjusted R-squared |
-0.011904 |
S.D. dependent var |
2.347946 |
| S.E. of regression |
2.361879 |
Akaike info criterion |
4.629264 |
| Sum squared resid |
546.6902 |
Schwarz criterion |
4.830278 |
| Log likelihood |
-237.3510 |
F-statistic |
0.823546 |
| Durbin-Watson stat |
0.094549 |
Prob(F-statistic) |
0.570083 |
从回归结果来看，M2Z各滞后期的系数逐步增加，表明当期货币供应量的变化对物价水平的影响要经过一段时间才能逐步显现。但各滞后期的系数的t统计量值不显著，因此还不能据此判断滞后期究竟有多长。为此，我们做滞后12个月的分布滞后模型的估计，结果见表7.7。
表7.7
| Dependent Variable: TBZS |
| Method: Least Squares |
| Date: 07/03/05 Time: 17:09 |
| Sample(adjusted): 1997:02 2005:05 |
| Included observations: 100 after adjusting endpoints |
| Variable |
Coefficient |
Std. Error |
t-Statistic |
Prob. |
| C |
98.35668 |
0.467897 |
210.2102 |
0.0000 |
| M2Z |
-0.167665 |
0.121743 |
-1.377203 |
0.1720 |
| M2Z(-1) |
-0.032065 |
0.111691 |
-0.287084 |
0.7747 |
| M2Z(-2) |
-0.000995 |
0.111464 |
-0.008925 |
0.9929 |
| M2Z(-3) |
0.004243 |
0.113815 |
0.037276 |
0.9704 |
| M2Z(-4) |
0.106581 |
0.112727 |
0.945480 |
0.3471 |
| M2Z(-5) |
0.043217 |
0.113161 |
0.381908 |
0.7035 |
| M2Z(-6) |
0.117581 |
0.118460 |
0.992575 |
0.3237 |
| M2Z(-7) |
0.140418 |
0.115571 |
1.214988 |
0.2277 |
| M2Z(-8) |
0.220875 |
0.114368 |
1.931271 |
0.0567 |
| M2Z(-9) |
0.140875 |
0.115354 |
1.221247 |
0.2253 |
| M2Z(-10) |
0.180497 |
0.115895 |
1.557410 |
0.1230 |
| M2Z(-11) |
0.246911 |
0.125543 |
1.966752 |
0.0524 |
| M2Z(-12) |
0.392359 |
0.130058 |
3.016798 |
0.0034 |
| R-squared |
0.317136 |
Mean dependent var |
100.7830 |
| Adjusted R-squared |
0.213913 |
S.D. dependent var |
1.890863 |
| S.E. of regression |
1.676469 |
Akaike info criterion |
4.000434 |
| Sum squared resid |
241.7072 |
Schwarz criterion |
4.365158 |
| Log likelihood |
-186.0217 |
F-statistic |
3.072325 |
| Durbin-Watson stat |
0.265335 |
Prob(F-statistic) |
0.000906 |
表7.7显示，从M2Z到M2Z(-11)，回归系数都不显著异于零，而M2Z（-12）的回归系数t统计量值为3.016798，在5％显著性水平下拒绝系数为零的原假设。这一结果表明，当期货币供应量变化对物价水平的影响在经过12个月（即一年）后明显地显现出来。为了考察货币供应量变化对物价水平影响的持续期，我们做滞后18个月的分布滞后模型的估计，结果见表7.8。
表7.8
| Dependent Variable: TBZS |
| Method: Least Squares |
| Date: 07/03/05 Time: 17:08 |
| Sample(adjusted): 1997:08 2005:05 |
| Included observations: 94 after adjusting endpoints |
| Variable |
Coefficient |
Std. Error |
t-Statistic |
Prob. |
| C |
97.41411 |
0.370000 |
263.2815 |
0.0000 |
| M2Z |
-0.083649 |
0.094529 |
-0.884900 |
0.3791 |
| M2Z(-1) |
-0.116744 |
0.093984 |
-1.242161 |
0.2181 |
| M2Z(-2) |
-0.119939 |
0.094428 |
-1.270156 |
0.2080 |
| M2Z(-3) |
-0.092993 |
0.095720 |
-0.971509 |
0.3345 |
| M2Z(-4) |
-0.032912 |
0.095823 |
-0.343468 |
0.7322 |
| M2Z(-5) |
-0.023891 |
0.097813 |
-0.244256 |
0.8077 |
| M2Z(-6) |
0.017290 |
0.100645 |
0.171794 |
0.8641 |
| M2Z(-7) |
0.028288 |
0.097570 |
0.289929 |
0.7727 |
| M2Z(-8) |
0.048708 |
0.095877 |
0.508021 |
0.6129 |
| M2Z(-9) |
0.025995 |
0.097569 |
0.266422 |
0.7907 |
| M2Z(-10) |
0.118247 |
0.096764 |
1.222011 |
0.2256 |
| M2Z(-11) |
0.157408 |
0.102558 |
1.534815 |
0.1291 |
| M2Z(-12) |
0.271281 |
0.112316 |
2.415326 |
0.0182 |
| M2Z(-13) |
0.325760 |
0.109217 |
2.982684 |
0.0039 |
| M2Z(-14) |
0.396242 |
0.107046 |
3.701601 |
0.0004 |
| M2Z(-15) |
0.335482 |
0.106776 |
3.141941 |
0.0024 |
| M2Z(-16) |
0.270811 |
0.107222 |
2.525697 |
0.0137 |
| M2Z(-17) |
0.200024 |
0.109278 |
1.830415 |
0.0712 |
| M2Z(-18) |
0.169696 |
0.101547 |
1.671114 |
0.0989 |
| R-squared |
0.610520 |
Mean dependent var |
100.6085 |
| Adjusted R-squared |
0.510519 |
S.D. dependent var |
1.795733 |
| S.E. of regression |
1.256348 |
Akaike info criterion |
3.480597 |
| Sum squared resid |
116.8024 |
Schwarz criterion |
4.021724 |
| Log likelihood |
-143.5881 |
F-statistic |
6.105105 |
| Durbin-Watson stat |
0.308938 |
Prob(F-statistic) |
0.000000 |
结果表明，从滞后12个月开始t统计量值显著，一直到滞后16个月为止，从滞后第17个月开始t值变得不显著；再从回归系数来看，从滞后11个月开始，货币供应量变化对物价水平的影响明显增加，再滞后14个月时达到最大，然后逐步下降。
通过上述一系列分析，我们可以做出这样的判断：在我国，货币供应量变化对物价水平的影响具有明显的滞后性，滞后期大约为一年，而且滞后影响具有持续性，持续的长度大约为半年，其影响力度先递增然后递减，滞后结构为
型。
当然，从上述回归结果也可以看出，回归方程的
不高，DW值也偏低，表明除了货币供应量外，还有其他因素影响物价变化；同时，过多的滞后变量也可能引起多重共线性问题。如果我们分析的重点是货币供应量变化对物价影响的滞后性，上述结果已能说明问题。如果要提高模型的预测精度，则可以考虑对模型进行改进。根据前面的分析可知，分布滞后模型可以用子回归模型来代替，因此我们估计如下子自回归模型：
在Eviews工作文档的方程设定窗口中，输入
TBZS C TBZS(-1)
估计结果见表7.9。
表7.9
| Dependent Variable: TBZS |
| Method: Least Squares |
| Date: 07/10/05 Time: 23:48 |
| Sample(adjusted): 1996:03 2005:05 |
| Included observations: 111 after adjusting endpoints |
| Variable |
Coefficient |
Std. Error |
t-Statistic |
Prob. |
| C |
5.348792 |
1.938684 |
2.758982 |
0.0068 |
| TBZS(-1) |
0.946670 |
0.019081 |
49.61371 |
0.0000 |
| R-squared |
0.957596 |
Mean dependent var |
101.4946 |
| Adjusted R-squared |
0.957207 |
S.D. dependent var |
2.828904 |
| S.E. of regression |
0.585200 |
Akaike info criterion |
1.784126 |
| Sum squared resid |
37.32798 |
Schwarz criterion |
1.832947 |
| Log likelihood |
-97.01900 |
F-statistic |
2461.520 |
| Durbin-Watson stat |
1.779257 |
Prob(F-statistic) |
0.000000 |