Building ARIMA Model for Forecasting Vietnam’s Outbound Tourists

Vietnam Tourism has become a focal industry of the country, accounted for 4% of the GDP. The sector has also contributed a considerable amount to the state budget. Vietnam has been rated as a fully potential destination for international tourists, however we have not taken advantage of such strengths. Therefore, forecasting tourist arrivals in coming years is very useful for managers and investors, who can base on this to perform their businesses better. The paper applied the Box-Jenkins method to build an appropriate ARIMA model for forecasting outbound tourist arrivals to Vietnam, using monthly data from Vietnam National Administration of Tourism. The findings show that among the trial estimated models, ARIMA(12,1,12) is the  most appropriate model. Based on this model, shorterm forecasts on outbound tourist arrivals to Vietnam in early months, 2011 has been done with errors varying from 14 to 27%. However, ARIMA model could be used for forecast, but it could not be optimal selection since autocorrelation and moving average in the model are assumed tobe linear.