A) Estime la ecuacin de regresin.
Dependent Variable: M
Method: Least Squares
Date: 09/24/16 Time: 21:56
Sample: 1991 2015
Included observations: 25
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
PBI
TC
-30332.00
0.331325
7.864756
3315.936
0.010615
19.39787
-9.147342
31.21239
0.405444
0.0000
0.0000
0.6891
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.980825
0.979082
5008.387
5.52E+08
-246.8473
562.6648
0.000000
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn creer.
Durbin-Watson stat
63836.83
34628.72
19.98778
20.13405
20.02835
1.215620
B) Realice unas pruebas estadsticas de significancia individual y mltiple de los parmetros del
modelo.
INDIVIDUAL: T-STUDENT
Wald Test:
Equation: EQ02
Test Statistic
t-statistic
F-statistic
Chi-square
Value
df
Probability
6.715693
45.10053
45.10053
13
(1, 13)
1
0.0000
0.0000
0.0000
Value
Std. Err.
0.423548
0.063068
Value
df
Probability
-71.88504
5167.459
5167.459
13
(1, 13)
1
0.0000
0.0000
0.0000
Value
Std. Err.
Null Hypothesis: C(2)=1
Null Hypothesis Summary:
Normalized Restriction (= 0)
-1 + C(2)
Restrictions are linear in coefficients.
Wald Test:
Equation: TEST
Test Statistic
t-statistic
F-statistic
Chi-square
Null Hypothesis: C(3)=1
Null Hypothesis Summary:
Normalized Restriction (= 0)
-1 + C(3)
-1.011476
0.014071
Restrictions are linear in coefficients.
MULTIPLE: PRUEVA DE F
Wald Test:
Equation: EQ02
Test Statistic
t-statistic
F-statistic
Chi-square
Value
df
Probability
5.891497
34.70974
34.70974
13
(1, 13)
1
0.0001
0.0001
0.0000
Value
Std. Err.
0.412072
0.069943
Null Hypothesis: C(2)+C(3)=1
Null Hypothesis Summary:
Normalized Restriction (= 0)
-1 + C(2) + C(3)
Restrictions are linear in coefficients.
RECHAZAMOS HIPOTESIS NULA.
C) Auto correlacin: segn la prueva de durbin-Watson: si hay una
presencia de auto correlacin bajo de 1.2
Durbin-Watson stat
1.215620
Solucin:
Primera trasformacin: LOGARITMOS (Por que se evala tasas
de crecimiento) para quitar el problema de dispersin y para
permitir evaluaciones porcentualmente.
Dependent Variable: LM
Method: Least Squares
Date: 09/25/16 Time: 18:05
Sample: 1991 2015
Included observations: 16
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
LPBI
LTC
-6.843036
1.423548
-0.011476
0.793143
0.063068
0.014071
-8.627751
22.57150
-0.815611
0.0000
0.0000
0.4294
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
0.979694
0.976570
0.086253
0.096714
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
10.76581
0.563495
-1.895707
-1.750847
Log likelihood
F-statistic
Prob(F-statistic)
18.16566
313.6069
0.000000
Hannan-Quinn criter.
Durbin-Watson stat
-1.888289
0.768177
Segunda trasformacin: Inclusin de nuevas variables.
Residuos:
RESID
.16
.12
.08
.04
.00
-.04
-.08
-.12
92
94
96
98
00
02
04
06
08
10
12
14
Residuales:
12.0
11.5
11.0
.16
10.5
.12
10.0
.08
.04
9.5
.00
-.04
-.08
-.12
91
92
93
94
95
96
97
Residual
98
99
00
Actual
01
02
Fitted
09
13
14
15
POR LO TANTO:An existe presencia de autocorrelacion.
TERCERA TRASFORMACION: ADICIONAMOS VARIABLES DUMMI,
para sacar el impacto negativo que tenemos en nuestros
residuos.( quitamos datos del ao 2002, des de 3l ao 2001 al
2009)
DUM
2
0
92
94
96
98
00
02
04
06
08
10
12
14
Hacemos nuestra nueva regresin:
Dependent Variable: LM
Method: Least Squares
Date: 09/25/16 Time: 19:22
Sample: 1991 2015
Included observations: 16
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
LPBI
LTC
DUM
-6.387606
1.394181
-0.040972
-0.176407
0.639586
0.050374
0.014716
0.058278
-9.987094
27.67667
-2.784130
-3.026996
0.0000
0.0000
0.0165
0.0105
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.988486
0.985607
0.067602
0.054840
22.70433
343.4015
0.000000
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
10.76581
0.563495
-2.338041
-2.144894
-2.328150
1.358044
seguimos observando autor correlacin.
Adicionaremos rezagos por los resultados obtenidos a si mismo para
saber si lo que ocurri aos anteriores tiene impacto este ao para
poder medir espectativas.
Hoy= resago y maana=hoy
.primer resago: lpbi(-1) y segundo resago: ltc(-1)
Resultado:
Dependent Variable: LM
Method: Least Squares
Date: 09/25/16 Time: 19:36
Sample (adjusted): 1992 2015
Included observations: 13 after adjustments
Variable
Coefficient
Std. Error
t-Statistic
Prob.
C
LPBI
LTC
LPBI(-1)
LTC(-1)
DUM
-5.247789
0.988192
-0.030491
0.318924
-0.028199
-0.242181
1.077048
0.714638
0.033646
0.698815
0.026622
0.108729
-4.872380
1.382787
-0.906243
0.456379
-1.059206
-2.227371
0.0018
0.2092
0.3949
0.6619
0.3247
0.0612
R-squared
Adjusted R-squared
S.E. of regression
Sum squared resid
Log likelihood
F-statistic
Prob(F-statistic)
0.986712
0.977221
0.075162
0.039546
19.22292
103.9602
0.000002
Mean dependent var
S.D. dependent var
Akaike info criterion
Schwarz criterion
Hannan-Quinn criter.
Durbin-Watson stat
10.71201
0.498004
-2.034296
-1.773550
-2.087891
2.142632
Hacemos nuevamente el anlisis:
R-SQUARED: 0.99
DURBIN-WATSON: 2.14
POR LO TANTO, LA INCORPORACION DE RESAGOS EN ESPECIAL DEL
TC DEL AO PASADO EXPLICA FUERTEMENTE EN EL TC DEL AO
ACUAL. NO EXISTE AUTORRELACION.
FORMULA: ls m c pbi tc
genr lm=log(m)
genr lpbi=log(pbi)
genr ltc=log(tc)
ls lm c lpbi ltc
smpl @all
series dum=0
ls lm c lpbi ltc dum
