Sentiment Regimes and Adverse Selection in Cryptocurrency Markets
Working Paper DAI-2510 | Dissensus AI
Using the Crypto Fear & Greed Index and Bitcoin daily data, we document that sentiment extremity predicts excess uncertainty beyond realized volatility. Extreme fear and extreme greed regimes exhibit significantly higher spreads than neutral periods---a phenomenon we term the "extremity premium." Extended validation on the full Fear & Greed history (February 2018--January 2026, N = 2,896) confirms the finding: within-volatility-quintile comparisons show a significant premium (p < 0.001, Cohen's d = 0.21), Granger causality from uncertainty to spreads is strong (F = 211), and placebo tests reject the null (p < 0.0001). The effect replicates on Ethereum and across 6 of 7 market cycles. However, the premium is sensitive to functional form: comprehensive regression controls absorb regime effects, while nonparametric stratification preserves them. We interpret this as evidence that sentiment extremity captures volatility-regime interactions not fully represented by parametric controls---consistent with, but not conclusively separable from, the F&G Index's embedded volatility component. An agent-based model reproduces the pattern qualitatively. The results suggest that intensity, not direction, drives uncertainty-linked liquidity withdrawal in cryptocurrency markets, though identification of "pure" sentiment effects from volatility remains an open challenge.
| Finding | Result |
|---|---|
| Extremity premium significance | p < 0.001, Cohen's d = 0.21 |
| Granger causality (uncertainty to spreads) | F = 211 |
| Replication across assets | Confirmed on Ethereum |
| Replication across market cycles | 6 of 7 cycles |
| Placebo tests | Reject null (p < 0.0001) |
extremity premium, sentiment regimes, adverse selection, market microstructure, cryptocurrency, agent-based modeling
sentiment-microstructure-abm/
├── paper/ # LaTeX source and PDF
│ ├── main.tex # Paper source
│ ├── main.pdf # Compiled paper
│ ├── references.bib # Bibliography
│ ├── figures/ # Paper figures
│ └── tables/ # Paper tables
├── agents/ # Agent implementations (Market Maker, Informed, Noise, Arbitrageur)
├── simulation/ # Mesa ABM environment and matching engine
├── signals/ # Signal processing
├── analysis/ # Statistical analysis scripts
├── data_ingestion/ # Binance + Reddit API clients
├── feature_engineering/ # Microstructure features and sentiment analyzer
├── monitoring/ # Dashboard and metrics
├── demo/ # Demo scripts
├── tests/ # Test suite
├── config/ # Configuration files
├── arxiv-submission/ # arXiv submission package
├── CITATION.cff
├── requirements.txt
└── LICENSE
@article{farzulla2026extremity,
author = {Farzulla, Murad},
title = {The Extremity Premium: Sentiment Regimes and Adverse Selection in Cryptocurrency Markets},
year = {2026},
journal = {Dissensus AI Working Paper DAI-2510},
doi = {10.5281/zenodo.17989810}
}- Murad Farzulla -- Dissensus AI & King's College London
- ORCID: 0009-0002-7164-8704
- Email: murad@dissensus.ai
Paper content: CC-BY-4.0