Salut

a compatibilité (acceptabilité) du modèle avec une prévision de TCPIB de 2.

6) Tester l’hétéroscédasticité et l’autocorrélation dans le modèle et en déduire les conséquences. On donne :

|Corrélogramme | |

|Sample: 1977 2010 | |

|Included observations: 34 | |

|Autocorrelation |Partial Correlation | |AC | PAC | Q-Stat | Prob |

| . |**. | | . |**. | |1 |0.273 |0.273 |2.7677 |0.096 |

| . | . | | . | . | |2 |0.031 |-0.048 |2.8035 |0.246 |

| . | . | | . | . | |3 |-0.002 |0.002 |2.8037 |0.423 |

| . *| . | | . *| . | |4 |-0.060 |-0.064 |2.9516 |0.566 |

| . *| . | | . *| . | |5 |-0.095 |-0.066 |3.3291 |0.649 |

| .**| . | | .**| . | |6 |-0.224 |-0.198 |5.5246 |0.478 |

|ARCH Test: | |

|F-statistic |0.109283 | Probability |0.896845 | |

|Obs*R-squared |0.239371 | Probability |0.887199 | |

| | | | | |

|Dependent Variable: RESID^2 | |

|Method: Least Squares | |

|Included observations: 32 after adjusting endpoints| |

|Variable |Coefficient |Std. Error |t-Statistic |Prob. |

|C |2.393379 |0.898011 |2.665200 |0.0124 |

|RESID^2(-1) |0.099432 |0.214387 |0.463798 |0.6463 |

|RESID^2(-2) |0.006591 |0.213757 |0.030833 |0.9756 |

|R-squared |0.007480 | Mean dependent var |2.630800 | |

|Adjusted R-squared |-0.060969 | S.D. dependent var |3.442403 | |

|S.E. of regression |3.545791 | Akaike info criterion |5.458459 | |

|Sum squared resid |364.6063 | Schwarz criterion |5.595872 | |

|Log likelihood |-84.33535 | F-statistic |0.109283 | |

|Durbin-Watson stat |1.760682 | Prob(F-statistic) |0.896845 | |