Modeling tourist arrivals in Sri Lanka

Tourism industry development becomes one of the key strategies for economic development of many countries during the few decades. This study intends to predict monthly tourist arrivals in Sri Lanka for the period of six years from 2015 January to 2020 December using Time series forecasting simulation method. For the development of the model 516 monthly observations are made after adjusted for seasonality using X-13ARIMA-SEATS monthly seasonal adjustment Method. The study uses 36 alternative ARIMA models as simulation to find the best model to predict monthly tourist arrivals in Sri Lanka. The Akaike Information Criterion (AIC) is used as selection criteria for the models which find that the ARIMA (5, 1, 3) as the best model among the alternative models. According to the model, the expected tourist arrivals in the month of December 2020 will be 178, 705 tourists in Sri Lanka.