Fear & Greed Index

The index combines several inputs into one weighted score: realized volatility relative to recent averages, market momentum and trading volume, social media sentiment and engagement, Bitcoin's dominance of total crypto market cap (rotation into or out of BTC), and search trend data. Each input is normalized and blended daily into a single 0–100 number.

• 0–24, Extreme Fear — the market is trading defensively; historically this zone has often preceded relief rallies, though not reliably enough to be a standalone timing signal.
• 25–49, Fear — cautious positioning, below-average risk appetite.
• 50–74, Greed — optimistic positioning, above-average risk appetite.
• 75–100, Extreme Greed — the market is trading euphorically; historically this zone has often preceded corrections, for the same reason extreme fear has often preceded rallies — sentiment extremes tend to mean-revert.

Treat Fear & Greed as a contrarian context gauge: its edge is in the extremes, where crowd emotion tends to overshoot, and it carries little signal in the wide middle. The most common mistake is using it as a standalone trigger — a low reading can persist through an entire downtrend. It is most powerful as a confirmation layer: when a valuation metric like the MVRV Z-Score says the market is cheap and sentiment is simultaneously in extreme fear, the two agree from different angles.

Sentiment sits in the momentum & sentiment family the Risk Model weighs — deliberately as confirmation, secondary to valuation and on-chain signals rather than a driver. That ordering is intentional: sentiment is fast and noisy, so it refines the read at the margins instead of setting it. See the published category weights on the Methodology page, or the live reading on the free Fear & Greed tool.

Fear & Greed is a contrarian sentiment gauge, not a valuation or cycle-position signal — it can sit in one zone for weeks during a strong trend without that trend reversing. It's most useful as one input among several (which is why Alphabit's Risk Model treats sentiment as confirmation, secondary to valuation and on-chain signals) rather than as a standalone buy or sell trigger. Its exact input weights are also set by the data provider and aren't fully transparent, so the score is best read as a directional mood gauge, not a precise measurement.