chap3 - ARMA Models(Chapter 3 Motivation For `most time...

This preview shows pages 1–4. Sign up to view the full content.

1 1 Motivation: For ‘most’ time series { X t } , Wold Decomposition { X t } is a linear TS X t = ψ j Z t-j , { Z t } ~ WN(0, σ 2 ), = ψ(Β) Z t , where ψ(Β) = 1 + ψ 1 B + ψ 1 B 2 + . . . We approximate ψ(Β) using a ratio of polynomials which leads us to the class of ARMA models. ARMA Models (Chapter 3) j = 0 2 Write ARMA equations as: φ(Β) X t = θ(Β) Z t where φ( z) and θ( z) are the polynomials φ( z) = 1 −φ 1 z - - φ p z p , θ( z) = 1 1 z + + θ q z q . DEFINTION: { X t } is an ARMA(p,q) process if { X t } is stationary and if for every t, X t −φ 1 X t-1 -- φ p X t-p = Z t 1 Z t-1 + + θ q Z t-q where { Z t } ~ WN(0, σ 2 ). 3.1 ARMA(p,q) Processes L L L L

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
2 3 Causality: Solution { X t } is causal if we can write X t = ψ j Z t-j , with | ψ j | < . { X t } causal φ( z ) = 0 for |z | < 1 ψ(Β) = ( see p.85) Existence and Uniqueness: The ARMA eqns φ(Β) X t = θ(Β) Z t have a stationary solution (which is unique) if and only if φ( z ) = 0 for |z| = 1 j = 0 j = 0 8 θ( B) φ (B) 4 Invertibility: Solution { X t } is invertible if we can write Z t = π j X t-j , with | π j | < . { X t } invertible θ( z ) = 0 for |z | < 1 π(Β) = ( see p.86) j = 0 j = 0 φ( B) θ (B)
3 5 Ex 3.1.1 (An ARMA(1,1) process). X t −.5 X t-1 = Z t + .4 Z t-1 , { Z t } ~ WN(0, σ 2 ), AR polynomial : φ( z)=1 .5z (zero at z=2, so process is causal).

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

Page1 / 11

chap3 - ARMA Models(Chapter 3 Motivation For `most time...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online