Econometria

Tomando en cuenta las variables tipo de cambio, exportaciones e importaciones se realizaran las pruebas vistas durante el curso de econometría financiera tomando al índice de tipo de cambio como la variable a explicar.

El presente trabajo pretende, de acuerdo a los contenidos temáticos en la materia de Econometría Financiera, desarrollar cada uno de los métodos utilizando las variables de Índice de tipo de cambio, exportaciones e importaciones.

Métodos a desarrollar:

Ad-hoc

Koyck

Almon

Variables instrumentales

Prueba de causalidad de Granger

Extrapolación simple

Cointegración Engle-Granger

ARIMA

AD HOC

ORIGINAL TC-IMPORTACIONES

Dependent Variable: TC

Method: Least Squares

Date: 04/15/14 Time: 17:51

Sample: 1991M01 2013M12

Included observations: 276

Variable Coefficient Std. Error t-Statistic Prob.

C 86.28197 1.528718 56.44074 0.0000

IMPO -0.000370 8.46E-05 -4.374272 0.0000

R-squared 0.065275 Mean dependent var 80.40825

Adjusted R-squared 0.061863 S.D. dependent var 12.53288

S.E. of regression 12.13903 Akaike info criterion 7.837948

Sum squared resid 40375.54 Schwarz criterion 7.864183

Log likelihood -1079.637 Hannan-Quinn criter. 7.848476

F-statistic 19.13425 Durbin-Watson stat 0.079823

Prob(F-statistic) 0.000017

}MODELO MODIFICADO CON REZAGOS TC-IMPORTACIONES

Dependent Variable: TC

Method: Least Squares

Date: 04/15/14 Time: 18:20

Sample (adjusted): 1996M01 2013M12

Included observations: 216 after adjustments

Variable Coefficient Std. Error t-Statistic Prob.

C 82.17996 1.944991 42.25211 0.0000

IMPO -0.001282 0.000271 -4.724038 0.0000

IMPO(-48) 0.000652 0.000291 2.242263 0.0260

IMPO(-60) 0.000875 0.000372 2.349725 0.0197

R-squared 0.803441 Mean dependent var 78.07832

Adjusted R-squared 0.090754 S.D. dependent var 10.04521

S.E. of regression 9.578552 Akaike info criterion 7.375275

Sum squared resid 19450.71 Schwarz criterion 7.437780

Log likelihood -792.5297 Hannan-Quinn criter. 7.400527

F-statistic 8.153209 Durbin-Watson stat 0.074438

Prob(F-statistic) 0.000037

Substituted Coefficients:

=========================

TC = 82.1799554461 - 0.00128217037582*IMPO + 0.000652267258885*IMPO(-48) + 0.000875106557127*IMPO(-60)

Se dice que no existe evidencia de multicolinealidad debido a que los valores de t estadístico son mayores a 2 en valores absolutos y el valor de R2 es mayor a 0.80.

ORIGINAL TC-EXPORTACIONES

Dependent Variable: TC

Method: Least Squares

Date: 04/15/14 Time: 18:30

Sample: 1991M01 2013M12

Included observations: 276

Variable Coefficient Std. Error t-Statistic Prob.

C 85.25398 1.493526 57.08237 0.0000

EXPO -0.000316 8.46E-05 -3.730743 0.0002

R-squared 0.048342 Mean dependent var 80.40825

Adjusted R-squared 0.044868 S.D. dependent var 12.53288

S.E. of regression 12.24849 Akaike info criterion 7.855902

Sum squared resid 41106.97 Schwarz criterion 7.882137

Log likelihood -1082.114 Hannan-Quinn criter. 7.866429

F-statistic 13.91844 Durbin-Watson stat 0.078794

Prob(F-statistic) 0.000232

MODELO MODIFICADO CON REZAGOS TC-EXPORTACIONES

Dependent Variable: TC

Method: Least Squares

Date: 04/15/14 Time: 18:49

Sample (adjusted): 1999M04 2013M12

Included observations: 177 after adjustments

Variable Coefficient Std. Error t-Statistic Prob.

C 66.88960 1.486863 44.98705 0.0000

EXPO -0.000312 0.000124 -2.525369 0.0125

EXPO(-15) -0.000399 0.000173 -2.312042 0.0220

EXPO(-99) 0.002261 0.000274 8.259152 0.0000

R-squared 0.873465 Mean dependent var 75.32275

Adjusted R-squared 0.566069 S.D. dependent var 7.840966

S.E. of regression 5.165117 Akaike info criterion 6.144072

Sum squared resid 4615.369 Schwarz criterion 6.215850

Log likelihood -539.7504 Hannan-Quinn criter. 6.173182

F-statistic 77.53133 Durbin-Watson stat 0.287755

Prob(F-statistic) 0.000000

Substituted Coefficients:

=========================

TC = 66.8896043467 - 0.000312413633977*EXPO - 0.000399239631086*EXPO(-15) + 0.0022614316336*EXPO(-99)

Se dice que no existe evidencia de multicolinealidad debido a que los valores de t estadístico son mayores a 2 en valores absolutos y el valor de R2 es mayor a 0.80.

KOYCK

Dependent Variable: TC

Method: Least Squares

Date: 04/15/14 Time: 20:13

Sample (adjusted): 1991M02 2013M12

Included observations: 275 after adjustments

Variable Coefficient Std. Error t-Statistic Prob.

C 3.272174 1.515207 2.159555 0.0317

IMPO -3.84E-06 2.45E-05 -0.156821 0.8755

TC(-1) 0.959140 0.016829 56.99247 0.0000

R-squared 0.927565 Mean dependent var 80.34499

Adjusted R-squared 0.927032 S.D. dependent var 12.51151

S.E. of regression 3.379680 Akaike info criterion 5.284288

Sum squared resid 3106.848 Schwarz criterion 5.323744

Log likelihood -723.5896 Hannan-Quinn criter. 5.300123

F-statistic 1741.538 Durbin-Watson stat 1.638265

Prob(F-statistic) 0.000000

ρ ̂=1-d/2=1-1.638265/2 = 0.1808675

h=ρ ̂√(n/(1-n(var〖 α〗^(2)) ))=

0.1808675√(276/(1-276(〖0.016829〗^(2)) )) =

=0.1808675 x 59.52917772 = 10.76689355

Criterio de decisión

95%----- <[1.96]= No hay autocorrelación

...