The Leader-Follower Structure of Crude Oil and Commodity Currencies: A Multi-Scale Lead-Lag Assessment of USOIL and USDCAD

This whitepaper extends a structured lead-lag validation framework to the oil-currency domain, focusing on the dynamic relationship between USOIL (Crude Oil) and USDCAD (Spot FX). Moving away from traditional methods that rely solely on quantile causality, this research applies event-based detection across multiple timeframes (M1, H1, D1) combined with layered statistical validation. The pipeline includes bootstrap resampling (10,000 iterations), binomial testing, permutation analysis with temporal-shift preservation, out-of-sample validation (70/30 chronological split), and Benjamini-Hochberg False Discovery Rate (FDR) correction. The empirical results demonstrate that the leader-follower structure between USOIL and USDCAD is highly episodic and context-dependent. At the intraday micro scale (M1) and long-term structural scale (D1), no significant lead-lag effects are detected, with outcomes dominated by noise and random synchronization. At the session baseline (H1), permutation tests suggest localized significance within 2–3 bars, but this effect weakens under FDR correction and fails to generalize out-of-sample. These findings align with recent literature (e.g., Ateba et al., 2024), confirming that no unconditional robust lead exists between oil price shocks and advanced oil-exporting exchange rates under normal market conditions. From a practical perspective, this study serves as a methodological blueprint for intermarket analysis, emphasizing that apparent lead-lag relationships must survive rigorous multi-layered statistical filters before being considered credible for quantitative trading strategies. This study constitutes an empirical extension of the core framework established in Wibowo (2026), "Lead-Lag Detection in Financial Markets: A Structured Framework Integrating Bootstrap, FDR, Permutation Tests, and Future AI Extensions" (DOI: 10.5281/zenodo.22911049).