Fear, greed and bitcoin return distributions: evidence from digital asset market states

Le résumé fourni par la source

Abstract Sentiment indicators are widely used in digital asset markets, but their economic meaning remains ambiguous. In the Bitcoin market, the Crypto Fear and Greed Index is often treated as a simple trading signal, although it may be more relevant for identifying market states and distributional risk. This study examines whether the index should be interpreted as a digital asset market-state variable rather than as a short-horizon buy-or-sell indicator. Using 2,945 daily observations from 1 February 2018 to 27 February 2026, we link the index and its five labelled states to one-day, seven-day and thirty-day forward Bitcoin returns. The results show that sentiment labels are persistent: Extreme Fear remains Extreme Fear on 82.0% of the next days, while Greed remains Greed on 80.0%. A one-standard-deviation increase in the index is associated with a 0.14%-point higher one-day and a 0.88%-point higher seven-day forward returns; a moving-block confidence interval based on 30-day blocks remains positive for the one-day estimate but includes zero for the seven-day estimate. Quantile point estimates are positive across several parts of the return distribution, although moving-block intervals show substantially lower precision than conventional quantile inference. Expanding-window forecasts do not outperform the historical mean benchmark. The study’s originality lies in separating state persistence, distributional relevance and real-time predictive usefulness. The findings suggest that sentiment dashboards may support risk interpretation, but should not be used as independent trading rules.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.