TimeSeriesBook.pdf

# 4 as a reminder the theoretical autocorrelation

• 435

This preview shows pages 100–103. Sign up to view the full content.

4 As a reminder: the theoretical autocorrelation coefficients are ρ ( h ) = φ | h | .

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

82 CHAPTER 4. ESTIMATION OF MEAN AND ACF 0 5 10 15 20 25 30 35 40 -1 -0.5 0 0.5 1 order correlation coefficient theoretical ACF estimated ACF upper bound for AR(1)-confidence interval lower bound for AR(1)-confidence interval Figure 4.3: Estimated autocorrelation function of an AR(1) process with φ = 0 . 8 and corresponding 95 percent confidence interval for T = 100 4.3 Estimation of the partial autocorrelation function According to its definition (see definition 3.2), the partial autocorrelation of order h , α ( h ), is equal to a h , the last element of the vector α h = Γ - 1 h γ h (1) = R - 1 h ρ h (1). Thus α h and consequently a h can be estimated by ˆ α h = b Γ - 1 h ˆ γ h (1) = ˆ R - 1 h ˆ ρ h (1). As ρ ( h ) can be consistently estimated and is asymptotically nor- mally distributed (see Section 4.2), the continuous mapping theorem (see Appendix C) ensures that the above estimator for α ( h ) is also consistent and asymptotically normal. In particular we have for an AR(p) process (Brock- well and Davis, 1991) T ˆ α ( h ) d ----→ N(0 , 1) for T → ∞ and h > p. This result allows to construct, as in the case of the autocorrelation coef- ficients, confidence intervals for the partial autocorrelations coefficients. The 95 percent confidence interval is given by ± 1 . 96 T . The AR(p) process is char- acterized by the fact that the partial autocorrelation coefficients are zero for h > p . ˆ α ( h ) should therefore be inside the confidence interval for h > p and outside for h p . Figure 4.4 confirms this for an AR(1) process with φ = 0 . 8.
4.4. ESTIMATION OF THE LONG-RUN VARIANCE 83 0 2 4 6 8 10 12 14 16 18 20 -1 -0.5 0 0.5 1 order partial autocorrelation coefficient estimated PACF lower bound for confidence interval upper bound for confidence interval Figure 4.4: Estimated PACF for an AR(1) process with φ = 0 . 8 and corre- sponding 95 percent confidence interval for T = 200 Figure 4.5 shows the estimated PACF for an MA(1) process with θ = 0 . 8. In conformity with the theory, the partial autocorrelation coefficients converge to zero. They do so in an oscillating manner because θ is positive (see formula in Section 3.5) 4.4 Estimation of the long-run variance For many applications 5 it is necessary to estimate the long-run variance J which is defined according to equation (4.1) as follows 6 J = X h = -∞ γ ( h ) = γ (0) + 2 X h =1 γ ( h ) = γ (0) 1 + 2 X h =1 ρ ( h ) ! . (4.5) This can, in principle, be done in two different ways. The first one con- sists in the estimation of an ARMA model which is then used to derive the implied covariances as explained in Section 2.4 which are then inserted into 5 For example, when testing the null hypothesis H 0 : μ = μ 0 in the case of serially correlated observations (see Section 4.1); for the Phillips-Perron unit-root test explained in Section 7.3.2.

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

This is the end of the preview. Sign up to access the rest of the document.
• Spring '17
• Raffaelle Giacomini

{[ snackBarMessage ]}

### What students are saying

• As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

Kiran Temple University Fox School of Business ‘17, Course Hero Intern

• I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

Dana University of Pennsylvania ‘17, Course Hero Intern

• The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

Jill Tulane University ‘16, Course Hero Intern