Comparative Revenue Forecasting of GST vs SST in Malaysia: Time Series and Regression Analysis
Zulhaimi Bin Daud
School of Quantitative Sciences, Universiti Utara Malaysia, 06010 UUM, Sintok, Kedah, Malaysia.
Muhammad Bin Mat Yusof *
School of Quantitative Sciences, Universiti Utara Malaysia, 06010 UUM, Sintok, Kedah, Malaysia.
*Author to whom correspondence should be addressed.
Abstract
The evolution of Malaysia’s indirect tax framework is defined by three major regimes consisting of the single-stage Sales and Service Tax (SST 1.0), the multi-stage Goods and Services Tax (GST) introduced in 2015, and the reinstated SST 2.0 in 2018. This study utilizes a methodological triangulation approach to quantify the long-term revenue implications and fiscal opportunity costs associated with these shifts. Holt-Winters and ARIMA models were applied to analyse long-run historical trends (SST 1.0 and GDP), while bivariate regression models were utilized to forecast the more recent, albeit data-constrained, GST and SST 2.0 series. Utilizing quarterly data, the analysis explicitly acknowledges limited sample sizes for the GST (n=14) and SST 2.0 (n=22) periods. Findings reveal a significant and widening fiscal gap, by Q4 2028, a continued GST regime is projected to generate RM21.8 billion quarterly, effectively double the RM10.9 billion projected yield of SST 2.0, representing a quarterly fiscal opportunity cost of RM10.9 billion. The study identifies a “Predictability-Potency Paradox,” where SST 2.0 exhibits greater budgetary predictability R² = .601), whereas GST demonstrates superior revenue potency and buoyancy (β = 0.048). These results underscore a critical “Policy Trilemma” for Malaysian policymakers, necessitating a complex trade-off between fiscal maximization, political acceptability, and administrative simplicity.
Keywords: Indirect tax, GST vs SST, revenue forecasting, time series analysis, regression analysis