A mood gauge, not a trading order
The index compresses several market observations into one score: 0 represents extreme fear and 100 extreme greed. It is useful because it gives a consistent name to the market’s emotional state. It is dangerous when that name is mistaken for an instruction to buy or sell.
What sits behind a number from 0 to 100
Alternative.me describes volatility and market momentum or volume as the largest inputs, followed by social media, Bitcoin dominance and search trends. Surveys are listed in the methodology but currently paused. The published index is focused on Bitcoin, so it should not be treated as a universal reading for every token.
Fear, greed and Bitcoin across the full sample
Weekly display points preserve the long-run shape; calculations use every matched daily observation.
How to read this chart
Read the upper scale
The coloured line is the published sentiment score: lower means more fear, higher means more greed.
Compare the lower line
The blue line is Bitcoin’s daily close on a logarithmic scale, so equal vertical distances represent comparable percentage moves.
Do not infer causality
The shared timeline shows when the observations coincided. It does not prove that sentiment caused the next price move.
Price data: Binance Spot · BTCUSDT · 1D · UTC. No exchange endorsement or trading link.
What happened after sentiment changed state
We matched each published daily index score to the BTCUSDT close on the same UTC date. Between 1 February 2018 and 19 September 2026, both values are available for 3,149 dates; four calendar dates in that span are not in the matched sample. The timeline shows that fear and greed often persist for several days.
To avoid counting the same mood repeatedly, the main test uses only the first day after the published classification changes. That produces 727 episodes: 128 entered Extreme Fear and 64 entered Extreme Greed. Returns are measured from that day’s close to the close 7, 30 and 90 calendar days later.
Extreme Fear did not behave like a dependable immediate bargain. Across 128 entries, the median return was −0.75% after 7 days, −1.63% after 30 days and −10.17% after 90 days. Only 46.1% of 30-day outcomes were positive. Panic can deepen before a durable recovery begins.
Extreme Greed was not a dependable instant sell signal either. Its 64 entries had a positive 7-day median of +0.96%, then negative medians of −2.02% after 30 days and −2.38% after 90 days. Momentum sometimes continued before later weakness appeared.
How to read this chart
Choose a horizon
Switch between 7, 30 and 90 days. The printed percentage is the median return after a new state began.
Use the centre line
Bars to the right are positive; bars to the left are negative. Length shows magnitude, not certainty.
Check the sample
The count is episodes, not days. Extreme states have fewer observations and wider uncertainty.
Why the popular slogan fails a simple test
If every day is counted, Extreme Greed looks much stronger: its 30-day median is +5.58%. That is not a contradiction. Long bullish runs generate many consecutive greedy observations, so repeated daily rows partly describe the same trend. The episode test asks the cleaner question: what happened when the state first changed?
| State | Episodes | 30-day median · all days | 30-day median · entries | positive outcomes |
|---|---|---|---|---|
| Extreme Fear | 128 | +1.61% | -1.63% | 46.1% |
| Fear | 217 | 0.00% | -0.46% | 47.9% |
| Neutral | 158 | +2.79% | +4.46% | 57.7% |
| Greed | 160 | +1.35% | +0.29% | 50.9% |
| Extreme Greed | 64 | +5.58% | -2.02% | 43.8% |
Forward windows overlap in the daily view. Episode entries reduce repeated counting but are not statistically independent market experiments.
Persistence, momentum and the problem of causality
The index is partly built from price-adjacent variables, including momentum and volatility. A high reading can therefore be the consequence of an advance as much as a cause of what follows. The 2026 academic study reached a similar practical conclusion: the index describes market state better than it forecasts returns by itself.
How to use the index while exercising your own judgement
The useful workflow is conditional. First identify the regime, then inspect trend, volatility, liquidity and leverage. An extreme reading becomes a prompt for questions: is momentum accelerating, are positions crowded, and does price confirm or reject the mood?
- Name the state, but do not act on the label alone.
- Confirm whether price is above or below its prevailing trend.
- Check volatility, liquidity and leverage before judging the reading.
- Define the holding horizon and invalidation point in advance.
- Record the decision so the index cannot become an after-the-fact story.
Verdict: context is useful; certainty is not
The evidence rejects both shortcuts. Fear is not automatically cheap, and greed is not automatically a top. The index adds context, but entry price, horizon, risk limits and the mechanism behind the move still decide the trade.
Is Extreme Fear a buy signal?
No. In this sample, first entries into Extreme Fear had negative median returns at 7, 30 and 90 days.
Is Extreme Greed a sell signal?
No. The 7-day median remained positive, while the 30- and 90-day medians later turned negative.
Why do daily readings and episodes differ?
Daily readings repeatedly count persistent regimes. Episodes count only the first day after the classification changes.
Research note
Historical distributions are not forecasts. Bitcoin can move sharply, and a composite sentiment index cannot replace position sizing or an independent risk plan.
Sources and methodology
Price data: Binance Spot · BTCUSDT · 1D · UTC. No exchange endorsement or trading link.