Hedging Bitcoin ETFs Risk: A Comparison of Static, Semi-Dynamic and Dynamic Approaches

Bitcoin futures ETFs represent a relatively recent and rapidly evolving financial innovation whose hedging efficiency remains largely unexplored in the existing literature. Addressing this gap, this study provides a comprehensive comparison of static, semi-dynamic and dynamic hedging strategies across three pairs of Bitcoin spot and futures ETFs—the most liquid pair (IBIT and BITO), the stable and aggressive barbelled pair (FBTC and BITS), and the multi-asset barbelled pair (ARKB and BTF) over the sample period from January 2024 to July 2026. Their daily adjusted price data and transformed returns are analysed using Johansen cointegration test, Engle's DCC model, and, static, semi-dynamic and dynamic optimal hedge ratio methodologies. Bitcoin spot ETFs demonstrate superior risk-return profiles relative to futures counterparts for two of the three pairs, as IBIT and ARKB exhibit higher mean returns than their paired futures ETFs, BITO and BTF respectively, whereas FBTC records a lower mean return than its futures counterpart, BITS. The futures-based BTF alone records a negative mean return alongside elevated volatility, reflecting persistently weaker performance among the futures counterparts. The estimates show that all pairs generate positive and statistically significant hedge ratios across static, semi-dynamic, and dynamic specifications, along with persistent volatility shock transmission between Bitcoin spot and futures ETFs returns across all pairs. These results confirm that investors holding long positions in Bitcoin spot ETFs should optimally hedge by simultaneously taking short positions in their paired futures ETFs. Static and semi-dynamic hedging strategies (Naive, OLS, VAR, GARCH, TARCH) consistently and significantly outperform the fully dynamic diagonal-VECH and diagonal-BEKK-GARCH specifications across all three ETF pairs, while Diagonal-BEKK-TARCH shows pair-dependent performance, confirming the robustness of simpler models over more heavily parameterized dynamic alternatives. Notably, a fundamental divergence emerges between statistical and economic performance, as both in-sample and out-sample variance-reduction and significance testing consistently favour static and semi-dynamic strategies for all three pairs, while the risk-return framework favours these strategies as the most effective strategies for two of the three pairs, with one pair showing a complete failure of risk-adjusted effectiveness despite strong variance reduction under every model tested. These findings underscore that optimal hedging strategy selection requires evaluation through both statistical and economic performance lenses facilitating objective-driven strategy selection for diverse market participants navigating the inherently volatile Bitcoin ETFs landscape marked by regulatory milestones and geopolitical shocks.