In: Statistics and Probability
Consider the model Yt = ΦYt−3 + et − θet−1, where et has variance σ 2 .
(a) Identify Yt as a certain SARIMA(p, d, q) × (P, D, Q)s model. That is, specify each of p, d, q, P, D, Q, and s. You may assume that Φ < 1.
(b) Find the variance of Yt .
(c) What are the forecasts for Yt+1 and Yt+4?
(d) What are the error variances for your forecasts above?
(e) If σ 2 = 1, Φ = .7, and θ = −.5, find 95% limits for your forecasts above. You may assume that et are normally distributed. Also, the four most recent values are yt−3 = 0.13, yt−2 = −0.50, yt−1 = 0.38, and yt = 1.53. Similarly, the four most recent et values are 0.08, −0.60, 0.75, and 0.95.
Answer a
SARIMA(p, d, q) × (P, D, Q)S with
Without differencing operations, the model could be written more formally as
Φ(BS)φ(B)(xt - μ) = Θ(BS)θ(B)wt
Now here we are having
=>
Hence here we are having SARIMA((1,3), 0, 1) × (0, 0, 0)0 that means it is an AMRA model with 1st and 3rd oreder of AR and MA(1)
Answer b
remaining cross product terms will be zero because of definition of Yt and et that et are independently distributed and observations and error terms are independent of each other.
=> by using variance of et
=>
Answer c
When t=t+1 then the equation would be
=>
When t=t+4 then the equation would be
=>
Answer d
When t=t+1,
were mean of et is zero
Now Variance of forecast error would be
Similarly hen t=t+4, and Variance of forecast error would be
Answer d
we are having σ 2 = 1, Φ = .7, and θ = −.5
a 95% prediction interval for the h-step forecast is
where σ^2h is an estimate of the standard deviation of the h-step forecast distribution.
Yt =0.7 Yt-3 + et+0.5 et-1 |
Confedence interval | |||||
time | Yt | et | Var(yt) | Lower bound | Upper bound |
t-3 | 0.13 | 0.08 | 0.84 | -1.5143 | 1.774295 |
t-2 | -0.50 | −0.60 | 0.84 | -2.1443 | 1.144295 |
t-1 | 0.38 | 0.75 | 0.84 | -1.2643 | 2.024295 |
t | 1.53 | 0.95 | 0.84 | -0.1143 | 3.174295 |