ESTIMATING THE OPTIMAL HEDGE RATIOS OF SHANGHAI CRUDE OIL FUTURES USING A DENOISING-MULTIFRACTAL DUAL INTELLIGENT INTEGRATION APPROACH

Crude oil markets are characterized by noise and multifractal features, which undermine the reliability of traditional hedge ratio estimation models. Therefore, this study develops a denoising-multifractal dual intelligent integration framework for estimating optimal hedge ratios in the Shanghai crude oil futures on the Shanghai International Energy Exchange (INE) and crude oil spot markets, with a primary focus on minimizing spot risk while enhancing hedging returns. The framework begins with an innovative use of the Complementary Ensemble Empirical Mode Decomposition (CEEMD) approach to remove high-frequency noise, improving data quality. It then employs the Multifractal Detrended Cross-Correlation Analysis (MF-DCCA) method to capture the multifractal structure of crude oil futures and spot markets. Building on these results, the Flower Pollination Algorithm (FPA) is employed to integrate hedge ratios in a two-stage manner. First, local hedge ratios are aggregated across time scales within each fluctuation level. Second, hedge ratios are further integrated across different fluctuation amplitudes. This unique design allows the model to fully exploit the multi-scale and multi-fluctuation information for deriving the optimal hedge ratios. Empirical analysis confirms the existence of significant noise and multifractal properties in crude oil markets. Moreover, the hedging results show that the proposed model outperforms all competing methods, achieving higher accumulated returns, Hedging Effectiveness (HE), and Sharpe ratios in most cases. The study offers an effective hedge ratio estimation method for investors in the INE crude oil futures and spot markets.