Suppose you are employed by a firm which markets interior trim insurance to new car owners. The insurance can only be provided on new cars and is valid for a 12-month period after which the policy lapses and cannot be renewed. The insurance covers things like stain removal from upholstery and minor repairs from scratches and other blemishes. You have been asked to predict the likely potential market for this insurance service for each month from July 2017 -Jun 2019. Currently the firm offers the same insurance policy type for all passenger vehicles (excluding trucks and utilities) but is considering offering separate insurance policies for normal passenger vehicles and SUV type vehicles. You have been assigned the task of analysing relevant time series and providing forecasts for the relevant period. Your group will also need to provide recommendations to management on possible outcomes associated with the proposed segmentation outlined above. In addition to providing forecasts, you are also required to outline relevant considerations due to changing economic circumstances and other environmental factors.
Part 1: Individual
(a) Provide a relevant time series which would will be useful in forecasting the number of motor vehicles in Australia for July 17 – Jun 19. Justify your choice of time series.
(i) Provide plots of the relevant time series for the last 10 years.
(ii) Provide comment on the nature of the time series plots.
(iii) Comment on what factors are likely to have influenced the nature of the time series? What circumstances may have produced certain characteristics of the time series? Will those circumstances apply to future periods?
(b) Apply appropriate methods to smoothe the time series and help identify the relevant underlying components of the time series. Provide numerical estimates for the underlying systematic components.
(c) Provide at least one appropriate time series model for the time series you have selected. Justify your choice of model. For this model provide forecasts for Jul-17 to Jun-19.
