Abstract: Forecasting tourism is one of the important areas that need to be explored, as tourism is in direct contact with society. Tourism in the region is closely related to its economy, culture and the environment. As such, it affects the economic levels of the region, for example, increasing foreign exchange in the country and creating employment opportunities. Therefore, it is very important to know how tourism will develop in the future, which depends mainly on future demand (tourist arrivals). Many publications on tourism forecasting have appeared during the past years. Although different forecasting techniques can be used, the major conclusions are that time series models are simplest and the least expensive. The purpose of our research is to predict foreign visitor arrivals in Indonesia, Malaysia, and Japan by using Holt-Winters Methods (Additive, Multiplicative and Extended Holt-Winters Method) and α-Sutte Indicator. Data for our research is comprised of foreign visitor arrivals in Indonesia, Malaysia, and Japan from January 2008 to November 2017. The data is divided into 2 parts, namely fitting data and testing data. Based on the results of all four forecasting methods, we conclude that the Extended Holt-Winters method is most suitable. At the end of the analysis, using the Extended Holt-Winters method, we calculate monthly forecasts of tourist arrivals for all three countries in 2018.  

Key words: tourism; forecasting; foreign visitor arrivals; Holt-Winters method; α-sutte indicator

JEL codes: C61, C32, R11, Z32