Univariate Modelling Using Periodic and Non-Periodic Analysis: Inbound Tourism to Japan, Australia and New Zealand Compared

This paper examines three hypotheses using the time-series between 1978 and 1995 of quarterly tourist flows into Japan, Australia and New Zealand. First, whether periodic models can be used to successfully forecast tourism arrival series. Second, whether the Basic Structural Model (BSM) can forecast as accurately as the Box Jenkins (ARIMA) model. Third, whether the integrated seasonal univariate models forecast more accurately than the naive process. The paper concludes that the periodic model does not improve forecast accuracy. The BSM model performs at least as well as the ARIMA model, and the naive process is not a substitute for seasonal integrated models.