Autoregresive Intergrated Movement with Conditionally Heteroskedastic Time Series Model in Forecasting Tourist Arrival Da Nang City, Vietnam

Although tourist arrival forecast plays an increasingly important role for investors and local government in countries where tourism is an advantage, there has been still a big difference between forecasted and actual figures. As a result, the reliability and validity of the forecasting are not convincible. One of the main reasons is the forecasting method. Previous forecasting methods such as regression analyses are based on the assumption that the variables are stationary but it is not true for tourism data, which are time series. When two non-stationary variables are regressed, the spurious regression has produced wrong outcomes, resulting in the difference between forecast and reality. The purpose of this study is to apply the time-series econometric models: Autoregresive intergrated moving average with conditionally heteroskedastic to forecast tourist arrivals in Danang based on the historic data 2004-2014.