ECON 3350/7350 Univariate Time Series – II
Eric Eisenstat
The University of Queensland
Tutorial 3
Eric Eisenstat
(School of Economics) ECON3350/7350 Week 3 1 / 3
Pure AR(p), MA(q) and ARMA(p,q)
Eric Eisenstat (School of Economics) ECON3350/7350 Week 3 2 / 3
Strategy to Fitting ARMA(p,q)
1 Identification
1 Plot the data and decide an appropriate process to begin with (is there a trend?).
2 Use the SACF and SPACF to suggest whether AR or MA terms are needed.
2 Estimation and Diagnostics
1 Estimate a model you think will be (more than) adequate.
2 Look at the SACF and SPACF for the residuals to suggest whether
more or fewer AR or MA terms are needed.
3 Repeat steps 3 and 4 to find the set of models that have no
autocorrelation in the residuals.
4 Reduce this set by choosing models with ¡°low¡± IC metrics (AIC, BIC).
3 Inference
1 Compare results across this set of models; interpret accordingly.
Eric Eisenstat (School of Economics) ECON3350/7350 Week 3 3 / 3