This paper studies short-horizon changes in Bitcoin implied volatility using publicly observable state variables, including on-chain metrics, macroeconomic indicators, sentiment measures, and price-based controls. Treating crypto options markets as a high-frequency laboratory, the analysis documents modest out-of-sample predictability in day-ahead implied volatility changes while highlighting a sharp distinction between statistical predictability and economic realizability. The empirical framework emphasizes identification, payoff mapping, and market microstructure frictions rather than trading performance. A deliberately conservative economic evaluation illustrates how execution costs and payoff mis-specification constrain the exploitation of volatility signals, consistent with limits-to-arbitrage interpretations. The findings underscore the importance of payoff-consistent evaluation frameworks when interpreting short-horizon predictability in implied volatility dynamics.
Anuj Pal (Fri,) studied this question.