THÈSE Estimation, validation et identification des modèles ARMA ...
THÈSE Estimation, validation et identification des modèles ARMA ...
THÈSE Estimation, validation et identification des modèles ARMA ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Chapitre 2. Estimating weak structural V<strong>ARMA</strong> models 70<br />
0.0 0.5 1.0 1.5<br />
0.0 0.4 0.8 1.2<br />
Strong V<strong>ARMA</strong>: estimator 2J −1<br />
(a) (b) (c)<br />
Estimates of diag(Ω)<br />
Weak V<strong>ARMA</strong>: estimator 2J −1<br />
(a) (b) (c)<br />
Estimates of diag(Ω)<br />
0.0 0.5 1.0 1.5<br />
0.0 0.4 0.8 1.2<br />
Strong V<strong>ARMA</strong>: sandwich estimator of Ω<br />
(a) (b) (c)<br />
Estimates of diag(Ω)<br />
Weak V<strong>ARMA</strong>: sandwich estimator of Ω<br />
(a) (b) (c)<br />
Estimates of diag(Ω)<br />
Figure 2.2 – Comparison of standard and modified estimates of the asymptotic variance Ω of the<br />
QMLE, on the simulated models presented in Figure 2.1. The diamond symbols represent the mean,<br />
over the N = 1,000 replications, of the standardized squared errors n{â1(2,2)−0.95} 2 for (a) (0.02<br />
2 in the strong and weak cases), n ˆb1(2,1)−2 for (b) (1.02 in the strong case and 1.01 in the strong<br />
2 case) and n ˆb1(2,2) for (c) (0.94 in the strong case and 0.43 in the weak case).