Forecasting the service tax revenue collected in Malaysia / Nur Eleena Hamdan

Service tax is one of the dutiable taxes collected in Malaysia. There are 25 classes of services that are being dutied such as health, credit/charge card, consultant, management, hotels and security control. Prior to 1st January 2011, the rate of service tax was 6 percent. The purpose of this study is to forecast service tax revenue collected in Malaysia. The objective of this study is to identify the type of time series component for service tax revenue, to evaluate service tax revenue by using Univariate Modelling and Box Jenkins method, and to identify the best forecast model for service tax revenue. The data available for this study was from January 2006 to October 2014, a total of 106 observations. The analysis under Univariate modeling that is tested are Single Exponential Modelling, Double Exponential Modelling, Holt’s Method, ARRES and Holt-Winter and Seasonality. The finding of this study will allows us to be able to determine the best method to forecast for the time series. There is trend and seasonal component detected in the series. By comparing the MSE’s value between Univariate Modelling and Box Jenkins Method, it is found that Holt-Winter and Seasonality has the smallest MSE’s value. Therefore it has been chosen as the best model to generate 5 months ahead service tax revenue