Tourist Arrivals in Sri Lanka: A Comparative Study of Holt- Winter’s versus Box- Jenkin’s Modeling Methods

Abstract

Tourism is one of the vastly growing and largest industries in the world. Contribution of tourism to Sri Lanka’s total foreign exchange earnings in 2016 amounted to 14.2%. After the civil strife in Sri Lanka, tourist arrivals continue to grow annually. Therefore, the post- conflict tourist arrivals were considered for this study. Forecasting is an essential analytical tool in tourism policy and planning. In all the regions at all times, there is no specific model that outperforms other models regularly. Therefore, the objective of this study is to compare Holt-Winter’s and Box- Jenkin’s methods of modeling the tourist arrivals and to recommend a better method to forecast the future tourist arrivals in Sri Lanka. Appropriate tests were applied in modeling exercises for both methods. The results demonstrate that, during June 2009 to June 2017, nearly 10.5 million of tourists had visited the island. Both models are adequate for forecasting tourist arrivals. However, based on the forecasting accuracy measures of the model, the Box-Jenkin’s method outperforms the Holt- Winter’s method. The Box- Jenkin’s model gives approximately 90% forecasting accuracy and therefore it is recommended to forecast the tourist arrivals in Sri Lanka. Accordingly, around 1.15 million of tourists arrived in Sri Lanka in the second half of the year 2017, and it is about 5.7% increase compared to the same period in 2016. Further, over 235,000 tourists arrived in December 2017, which was the highest monthly arrival in the history of Sri Lanka tourism so far.