Question

Retail trade accounts form a significant amount of the Australian economy. It is always important to study the behaviour of retail turnovers and forecast future Australian retail trade. We have obtained a sample of quarterly Australian retail turnovers from Q1, 2006 to Q2, 2019.

Turnover of Australian retail trade (2006-2019) ($m) 100000 90000 80000 Turnover ($m) 70000 60000 Алллллллл. 50000 40000 о 4 12 16 20 24 28 32 36 40 44 40 52 Quarters a. “The time series plot above can be described by all 4 components: trend, seasonal, cyclic and irregular”. Do you agree with this statement? Explain why or why not.

Answer 1

(a) There is a trend with the seasonal and cyclic patterns. But the pattern is not irregular.

(b)

1. The time series plot shows seasonality.
2. A cyclic pattern is formed.

(c) The retail trade turnover for December is 5% more than average.

(d) Since MSE for Model A is less, we can say that Model A is the better forecast model.

Actual	Model A	Model B	Model A Error²	Model B Error²
80300	82900	81000	6760000	490000
92900	83500	93400	10240000	1.72E+08
81900	84100	79900	14440000	160000
		MSE	10480000	57420000