FORECASTING THE TREND OF PRIVATE CAR REGISTRATIONS IN THAILAND: A TIME SERIES ANALYSIS

Abstract

This research aimed to (1) analyze trends in private passenger vehicle registrations in Thailand, (2) compare the forecasting accuracy of multiple time series models, and (3) propose policy recommendations based on the forecasting results. A quantitative time series analysis approach was employed. The study utilized secondary monthly data on private passenger vehicle registrations from 2020 to 2024 (60 observations) obtained from the Department of Land Transport, Thailand, processed via Minitab software. Forecasting models evaluated included Linear Trend, Exponential Growth, Simple Exponential Smoothing, Quadratic Trend, Multiplicative Seasonal, and Winters’ Multiplicative models, evaluated by Mean Absolute Percentage Error (MAPE) and Mean Absolute Deviation (MAD).
The findings indicated that (1) private passenger vehicle registrations were projected to decline by 2–3% annually during 2025–2029 due to high household debt, rising interest rates, changing consumer preferences among younger generations, and government policy support for electric vehicles (EVs); (2) the Quadratic Trend Model demonstrated the highest forecasting accuracy with the lowest error rates (MAPE = 11.0, MAD = 24,426), proving its effectiveness in capturing nonlinear trends and structural market shifts; and (3) the results provided empirical evidence for decision-making in automotive, energy, and environmental sectors, emphasizing EV infrastructure, tax restructuring, and sustainable mobility transitions.