Forecasting and Structural Analysis of Vietnam's Import-Export ofGoods and Services, 2015-2024: A Multi-Model Time-Series Approach

Tran Ngoc Dan

Abstract

This study analyzes and forecasts the structure and trends of Vietnam's imports and exports of goods and services during 2015-2024, with an emphasis on industry, economic sector, SITC group, country/territory markets, major commodity groups, trade-to-GDP ratio, and services trade. Method: The study uses descriptive statistics, structural comparison, time-series visualization, and six Minitab forecasting models: Trend Analysis, Decomposition, Moving Average, Single Exponential Smoothing, Double Exponential Smoothing, and Winter's Method. MAPE is used as the main selection criterion; if MAPE is equal, MSD is used as a tie-breaker. Dataset: The dataset consists of official-style annual trade tables ID277-ID291, including total import-export turnover, export and import value by sector, SITC classification, partner markets, major commodities, trade-to-GDP ratio, and services trade. Results: For total import-export turnover (ID277), Trend Analysis is optimal with MAPE = 3, MAD = 9,345, and MSD = 145,029,092. For exports by market (ID282), Decomposition is optimal with MAPE = 4 and MSD = 256,030,069. For major exports (ID283), Winter's Method is optimal with MAPE = 97 and MSD = 9,087,557. For major imports (ID289), Single Exponential Smoothing is optimal with MAPE = 35 and MSD = 3,374,986. Value: The research provides a quantitative framework for identifying export-import dynamics, trade concentration risks, FDI dependence, and model-based policy implications for Vietnam's trade strategy.

Source: semanticscholar · PDF

Read the AI summary and key takeaways for traders on WOBR Quant Research.