Forecasting International Tourist Arrivals in Malaysia using Holt-Winters Model
Abstract
Malaysia's tourism sector has been significantly impacted by the COVID-19 pandemic, especially in terms of foreign visitor arrivals. The aim of this research is to forecast the future number of international tourist arrivals using the Holt-Winters additive method. Additionally, this study analyses the current trends of international tourist arrivals in Malaysia. The dataset employed spans from January 2012 to June 2023, encompassing periods before, during, and after the COVID-19 pandemic. The estimation of the Holt-Winters model is based on 126 observations (January 2012 until June 2022), while the validation of the model is based on 12 observations (from July 2022 to June 2023). Based on the estimated Holt-Winters model, the forecast value of monthly number of international tourist arrivals in Malaysia was generated for July 2023 to June 2024. The trend analysis shows that the number of Malaysia’s tourist arrivals fluctuated significantly from one month to another from January 2012 to January 2020. Since then, the number of tourist arrivals progressively declined to its lowest level starting in April 2020 during the implementation of the travel restrictions policy due to the COVID-19 pandemic and the trend remained constant until March 2022. The number of tourist arrivals then slowly increased in May 2022 until now. As it turned out, the forecast generated for the next 12 months is closer in reaching Malaysia Tourism's target of 27.3 million tourists by the end of 2024. The results demonstrate that, despite short-term fluctuations, the overall trend indicates a gradual increase in the number of international tourist arrivals to Malaysia between July 2023 to June 2024. This indicates that the trend of international tourist arrivals in Malaysia is returning to the normal pattern observed prior to the COVID-19 pandemic.
Copyright (c) 2025 Gading Journal for the Social Sciences (e-ISSN 2600-7568)

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Authors who publish with this journal agree to the following terms:
1. Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgement of the work's authorship and initial publication in this journal.
2. Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgement of its initial publication in this journal.
3. Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.


