From noise to macro fear: the emergence of predictability in bitcoin volatility

Purpose This paper investigates the regime shift in Bitcoin volatility dynamics. We test whether the market transitioned from a noise-driven speculative asset to a macro-sensitive instrument, disentangling the impact of the May 2021 Chinese regulatory crackdown from the 2022 global inflation cycle. Design/methodology/approach Using daily data from 2017 to 2025, we employ a Heterogeneous Autoregressive (HAR-RV) model to forecast log-volatility. To address confounding macro factors, we utilize the Quandt-Andrews (Sup-Wald) test for endogenous structural break detection, together with alternative volatility estimators (Rogers-Satchell and Parkinson). Findings We document a profound shift from a noisy, unpredictable market to a highly predictable, macro-sensitive one. Prior to 2022, Bitcoin volatility was largely immune to external attention signals, functioning as a self-referential system where even retail hype (Google Trends) failed to predict volatility. We identify a three-phase maturation process. Following an endogenous structural break in August 2022 – aligning with peak global inflation – the market's predictability surged. Information-seeking regarding “Inflation” (Wikipedia) emerged as the dominant predictor, improving forecast accuracy by R2O OS ≈ 9.1%. Practical implications Our findings suggest that Bitcoin's current market maturity is a composite result of institutional restructuring (the 2021 China ban suppressing noise) and severe monetary anxiety (the 2022 inflation cycle introducing macro predictability). For risk managers, Wikipedia searches for “Inflation” now serve as a robust “fear gauge” for downside crash risk. Originality/value This study empirically disentangles regulatory shocks from macro-cycles, providing evidence that Bitcoin's current market maturity is a composite result of institutional restructuring and monetary anxiety.