Information flow and statistical coupling among Philippine oil prices, 1992–2026

Abstract

We characterize the statistical coupling among Philippine oil prices using monthly data from January 1992 to March 2026 (N = 411). Pearson correlation, mutual information, and transfer entropy are computed pairwise across four refined products (Ron 91, kerosene, avturbo, LPG), the Dubai crude benchmark, and the USD–PHP exchange rate. The refined products and Dubai crude form a tightly coupled cluster (ρ = 0.70–0.94 within products, 0.84–0.90 with crude), while the exchange rate sits apart (ρ = 0.40–0.62). Transfer entropy resolves the direction: Dubai crude is the dominant informational source (TE = 0.09–0.10), with reverse flows an order of magnitude smaller.