Assignment 4
STAT317/ECON323
Due 5pm Friday, 22 October 2021
A reminder that graphs help in interpretation and explanation and you are expected to present them properly.
1 Question 1 – 12 Marks
For assignment 3 I have supplied a csv file (3XX Q1 data.csv). This has four ARMA(p,q) time series with p = 0, 1, 2 q = 0, 1, 2. Do not worry about dates this time.
2
a b
c d
For each series decide the p and q values
Once you have decided on the p and q for each series fit that ARMA model.
Write your fitted ARMA model with your parameter estimates. Plot the modelled values on the same graph as the actual values
Question 2 – 13 marks
For this question use the time series you selected for the previous assignments. Use the data from the years 2000-2019 only i.e. suppress COVID effects.
a Plot your series
b Take the first differences and plot that series
c For the time series of first differences decide what p and q are. Ignore seasonal P and Q at this stage
d Fit that model to the series and write your fitted ARIMA(p,d,q) model with parameter estimates.
e Are your residuals white noise? Justify your answer.
f Now fit a ARIMA(p,d,q)×(0,1,1)12 model using the same p,d,q.
g Is this model better? Justify your reasoning. 1