FORECASTING OIL PRODUCTION USING SSA AND TREND REGRESSION IN THE WORLD’S TOP THREE OIL PRODUCERS

Abstract

The world's primary energy source, which plays a significant role in various global sectors, is petroleum. Due to geological influences, energy policies, and geopolitical factors, oil production patterns are often fluctuating and non-stationary. This also applies to the world's of top three oil-producing countries, that is, the United States, Saudi Arabia, and Iraq. This situation poses a challenge in developing accurate forecasting models to support global energy planning. This study aims to forecast oil production trends in the United States, Saudi Arabia, and Iraq using the Singular Spectrum Analysis (SSA) method combined with Trend Regression to obtain a forecasting model capable of capturing long-term patterns and mitigating the influence of short-term fluctuations. Annual oil production data for the period 1936–2024 were taken from Our World in Data. The analysis stages include data decomposition using SSA to separate trends, noise, and cycles, followed by trend component modeling using trend regression. Model evaluation was carried out using the coefficient of determination (R²). The results of the study indicate that the SSA and Trend Regression methods are able to produce stable and accurate projections, with the highest R² value in Saudi Arabia (0.98), followed by Iraq (0.75), and the United States (0.48). All three countries show an increasing production trend until 2034 with different patterns. The SSA and Trend Regression methods are effective in capturing the complex and non-stationary dynamics of oil production. This study provides both academic and practical contributions in the application of the SSA–Trend Regression hybrid method for global oil production forecasting as well as practical contributions for policymakers in projecting global oil production trends.