Real options, energy uncertainty, and nonlinear forecasting of Saudi oil consumption and exports

This study investigates how multidimensional energy uncertainty predicts Saudi Arabia’s direct crude-oil consumption (DUC) and crude-oil exports (ECO) using a nonlinear forecasting framework based on Hammerstein models optimized by Particle Swarm Optimization (PSO). Using monthly data from 2009–2023, the analysis includes five complementary uncertainty indicators, Oil Price Uncertainty (OPU), the Global Energy Uncertainty Index (EUI), the conditional variance of Brent returns (TGARCH-based), the Petroleum Markets EMV Tracker, and the OVX implied-volatility index. A strict real-time forecasting design is applied, with 2009M01–2019M06 used for estimation and 2019M07–2023M12 as a fully reserved out-of-sample evaluation period. Findings provide evidence of the superiority of nonlinear Hammerstein-PSO models over linear ARIMAX and SARIMAX standards, with the Dead-Zone specification showing the best performance. The EMV index appears to dominate other uncertainty indicators in forecasting local crude oil use, while the OVX index provides the most accurate forecasts for crude oil exports. The estimated nonlinear thresholds show distinct behavioural regimes consistent with real options theory. Findings show that local oil consumption is limited to responding only when oil-related uncertainty exceeds a critical EMV level, while adjustments in exports primarily react to severe spikes in global volatility. Aggregated uncertainty measures do not outperform individual indicators, suggesting that composite indicators may weaken information when uncertainty channels operate heterogeneously. The political implications indicate the value of uncertainty-based decision thresholds in improving strategies, export scheduling, hedging practices, and OPEC+ coordination. Overall, the results show that nonlinear models enhanced by Swarm intelligence provide a robust and interpretable framework for forecasting crude oil dynamics under uncertainty and offer actionable insights for energy planning in Saudi Arabia.