17.08.2013 Views

THÈSE Estimation, validation et identification des modèles ARMA ...

THÈSE Estimation, validation et identification des modèles ARMA ...

THÈSE Estimation, validation et identification des modèles ARMA ...

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Chapitre 2. Estimating weak structural V<strong>ARMA</strong> models 82<br />

✷<br />

The proof of Theorem 2.1 is compl<strong>et</strong>ed by the arguments of Wald (1949). More<br />

precisely, the compact s<strong>et</strong> Θ is covered by a neighborhood V(ϑ0) of ϑ0 and a finite<br />

number of neighborhoods V(ϑ1),...,V(ϑk) satisfying (2.31) with ϑ replaced by ϑi,<br />

i = 1,...,k. In view of (2.31) and (2.32), we have almost surely<br />

inf ˜ℓn(ϑ) = min<br />

ϑ∈Θ<br />

inf<br />

i=0,1,...,k ϑ∈V(ϑi)∩Θ<br />

˜ℓn(ϑ) = inf ˜ℓn(ϑ)<br />

ϑ∈V(ϑ0)∩Θ<br />

for n large enough. Since the neighborhood V(ϑ0) can be chosen arbitrarily small, the<br />

conclusion follows.<br />

2.9.3 D<strong>et</strong>ails on the proof of Theorem 2.2<br />

Lemma 9. Under the assumptions of Theorem 2.2, we have<br />

<br />

√ <br />

∂<br />

nsup <br />

<br />

˜ <br />

ℓn(ϑ) ∂ℓn(ϑ)<br />

<br />

<br />

− = oP(1).<br />

∂ϑ ∂ϑ <br />

ϑ∈Θ<br />

Proof of Lemma 9 : Similar to (2.30), Assumption A1 entails that, for k =<br />

1,...,k0,<br />

∞ ∂<strong>et</strong>(ϑ)<br />

= C<br />

∂ϑk i=1<br />

(k) ∂˜<strong>et</strong>(ϑ) t−1<br />

i (ϑ)Xt−i, = C<br />

∂ϑk i=1<br />

(k)<br />

i (ϑ)Xt−i, (2.33)<br />

C (k)<br />

i (ϑ) = ∂Ci(ϑ)<br />

<br />

<br />

, supC<br />

∂ϑk ϑ∈Θ<br />

(k)<br />

<br />

<br />

i (ϑ) ≤ Kρ i . (2.34)<br />

Similar to Lemma 6, we then have<br />

<br />

<br />

sup<br />

∂˜<strong>et</strong>(ϑ) ∂<strong>et</strong>(ϑ)<br />

−<br />

∂ϑ ′ ∂ϑ ′<br />

<br />

<br />

<br />

≤ Kρt . (2.35)<br />

Using (2.19), we have<br />

with<br />

a1 = 2<br />

√ n<br />

a2 = Tr<br />

√ n<br />

ϑ∈Θ<br />

n<br />

′ ∂e t(ϑ)<br />

t=1<br />

<br />

Σ −1<br />

e<br />

∂ϑk<br />

<br />

∂ ˜ ℓn(ϑ)<br />

∂ϑk<br />

− ∂ℓn(ϑ)<br />

<br />

= a1 +a2,<br />

∂ϑk<br />

− ∂˜e′ <br />

t(ϑ)<br />

Σ<br />

∂ϑk<br />

−1<br />

e <strong>et</strong>(ϑ)+ ∂˜e′ t(ϑ)<br />

Σ<br />

∂ϑk<br />

−1<br />

<br />

e (<strong>et</strong>(ϑ)− ˜<strong>et</strong>(ϑ))<br />

<br />

∂Σe<br />

.<br />

{˜<strong>et</strong>(ϑ)−<strong>et</strong>(ϑ)}˜e ′ t (ϑ)+<strong>et</strong>(ϑ){˜<strong>et</strong>(ϑ)−<strong>et</strong>(ϑ)} ′ Σ −1<br />

e<br />

∂ϑk

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!