Beyond static models - multiscale and time-varying approaches to hedging commodities

This study examines the effectiveness of various hedge ratio estimation techniques in the Indian derivatives market, focusing on three actively traded commodity futures: Zinc, Kapas, and Crude Oil. It compares the performance of static, multiscale, and time-varying hedge ratio models using methods such as Ordinary Least Squares (OLS), Dynamic Conditional Correlation–Generalized Autoregressive Conditional Heteroskedasticity (DCC-GARCH), and wavelet decomposition. The findings suggest that multiscale and time-varying models – especially the DCC-GARCH approach – offer more efficient hedging than static models. While the combined Wavelet-DCC (W-DCC) model performs well at lower time scales, the DCC-GARCH model is more robust and consistently delivers superior hedging performance across different time horizons.