You are on page 1of 6

Identificacin del Sistema de Ecuaciones simultaneas

C t + I t = Yt ...................................................................(1)

C t = 0 + 1Yt + e .........................................................(2)

Dependent Variable: C_INT


Method: Least Squares
Date: 06/13/07 Time: 15:21
Sample: 1950 2006
Included observations: 57
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
Y

2008.865
0.784735

780.1663
0.009006

2.574919
87.13131

0.0127
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

0.992808
0.992677
2372.139
3.09E+08
-522.8397
0.531069

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)

64229.37
27719.66
18.41543
18.48712
7591.865
0.000000

Los coeficientes de la regresin anterior son sesgados e incosistentes.


Para corregir el problema se debe obtener estos parmetros por otros mtodos de
estimacin diferentes al de mnimos cuadros ordinarios.
Problema de identificacin
Ecuaciones Reducidas
Reemplazando (2) en (1)
Yt = 0 + 1Yt + et + I t

Yt 1Yt = 0 + I t + et

0
1
1
+
It +
et ..................................(3)
1 1 1 1
1 1
Yt = 10 + 11 I t + wt ......................................................(3.1)
Yt =

Reemplazando (3) en (2)


1
1
Ct = 0 + 1 0 +
It +
et + et
1 1
1 1 1 1
Ct = 0 +

1 0
1
1
+
It +
et + et
1 1 1 1
1 1

0
1
1
+
It +
et ..................................(4)
1 1 1 1
1 1
C t = 20 + 21 I t + wt ......................................................(4.1)
Ct =

Las ecuaciones 3.1 y 4.1 son las ecuaciones reducidas.


Se tiene 3 ecuaciones:
10 = 20 =

0
= 0 11
1 1

11 =

1
1 1

21 =

1
= 1 11
1 1

Para encontrar 2 incgnitas

1 ; 2

En este sentido existen 3 ecuaciones para encontrar 2 incgnitas, en este sentido el sistema
estara sobre identificado.

Identificacin
G = Nmero de variables endgenas del
sistema
g = Nmero de variables endgenas de la

K = Nmero de variables exgenas del sistema


+ nmero de coeficientes independientes
k = Nmero de variables exgenas de la
ecuacin + nmero de coeficientes
independientes
R = Nmero de restricciones de la ecuacin

1era ecuacin
2da ecuacin
1+1=2

2
2

1era Ecuacin
2da Ecuacin

1
1

1era Ecuacin
2da Ecuacin

1
1

Condicin de Rango
1era Ecuacin
2da Ecuacin

R
1
1

G-1
1
1

R>=<G-1
Igual : Plenamente identificada
Igual : Plenamente identificada

Condicin de orden
1era Ecuacin
2da Ecuacin

Mnimos cuadrados indirectos

K-k
1
1

g-1
1
1

K-k >=<g-1
Igual : Plenamente identificada
Igual : Plenamente identificada

1er.- Se obtiene las ecuaciones reducidas


2da.- Se regresiona las ecuaciones reducidas por MCO
3ro.- Se obtienen los coeficientes de las ecuaciones reducidas
4to.Con estas resultados se calcula los valores de los parmetros de las ecuaciones
estructurales.
Dependent Variable: Y
Method: Least Squares
Included observations: 57

Sample: 1950 2006

Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
I_INT

15475.66
4.237477

3011.862
0.177290

5.138238
23.90144

0.0000
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

0.912180
0.910583
10524.64
6.09E+09
-607.7656
0.443327

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)

Dependent Variable: C_INT


Method: Least Squares
Included observations: 57

79288.53
35196.32
21.39528
21.46697
571.2789
0.000000

Sample: 1950 2006

Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
I_INT

15475.66
3.237477

3011.862
0.177290

5.138238
18.26095

0.0000
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

0.858416
0.855842
10524.64
6.09E+09
-607.7656
0.443327

10 = 20 = 15475.66 =

0
= 0 11 .........................(5)
1 1

11 = 4.237477 =

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)

1
..............................................(6)
1 1

de (6) obtenemos 1 = 0.76401052

1
= 1 11 .....................................................(7)
1 1
Reemplazando (6) en (5), obtenemos 2 = 3652.09298
21 =

Variables Instrumentales

64229.37
27719.66
21.39528
21.46697
333.4624
0.000000

1er.2da.3ro.4to.-

Se obtiene las ecuaciones reducidas


Se regresiona las ecuaciones reducidas por MCO
Se calcula la variable estimada de la ecuacin reducida
Se reemplaza la variable estimada en la ecuacin estructural.

Dependent Variable: Y
Method: Least Squares
Date: 06/13/07 Time: 15:11
Sample: 1950 2006
Included observations: 57
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
I_INT

15475.66
4.237477

3011.862
0.177290

5.138238
23.90144

0.0000
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

0.912180
0.910583
10524.64
6.09E+09
-607.7656
0.443327

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)

79288.53
35196.32
21.39528
21.46697
571.2789
0.000000

Yf=15478.66+4.237477 I_int
Dependent Variable: C_INT
Method: Least Squares
Date: 06/13/07 Time: 16:16
Sample: 1950 2006
Included observations: 57
Variable

Coefficient

Std. Error

t-Statistic

Prob.

C
YF

3652.093
0.764011

3598.313
0.041838

1.014946
18.26095

0.3146
0.0000

R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
Durbin-Watson stat

0.858416
0.855842
10524.64
6.09E+09
-607.7656
0.443327

Mnimos cuadrados en dos etapas

Mean dependent var


S.D. dependent var
Akaike info criterion
Schwarz criterion
F-statistic
Prob(F-statistic)

64229.37
27719.66
21.39528
21.46697
333.4624
0.000000

Se regresiona utilizando Two-Stage Least Squares y se incorpora las variables instrumento


Dependent Variable: C_INT
Method: Two-Stage Least Squares
Date: 06/13/07 Time: 16:23
Sample: 1950 2006
Included observations: 57
Instrument list: I_INT C
Variable

Coefficient

Std. Error

t-Statistic

Prob.

Y
C

0.764011
3652.093

0.009873
849.1640

77.38038
4.300811

0.0000
0.0001

R-squared
Adjusted R-squared
S.E. of regression
F-statistic
Prob(F-statistic)

0.992115
0.991972
2483.705
5987.723
0.000000

Mean dependent var


S.D. dependent var
Sum squared resid
Durbin-Watson stat

64229.37
27719.66
3.39E+08
0.443327

Mtodo Sistema de Ecuaciones


System: UNTITLED
Estimation Method: Two-Stage Least Squares
Date: 06/13/07 Time: 16:27
Sample: 1950 2006
Included observations: 57
Total system (balanced) observations 57

C(1)
C(2)

Coefficient

Std. Error

t-Statistic

Prob.

3652.093
0.764011

849.1640
0.009873

4.300811
77.38038

0.0001
0.0000

Determinant residual covariance


Equation: C_INT = C(1)+C(2)*Y
Instruments: I_INT C
Observations: 57
R-squared
0.992115
Adjusted R-squared
0.991972
S.E. of regression
2483.705
Durbin-Watson stat
0.443327

5952342.

Mean dependent var


S.D. dependent var
Sum squared resid

64229.37
27719.66
3.39E+08

You might also like