The number you see on our Fear & Greed page is not a poll and not a price prediction. It is a weighted mix of six data streams, published once a day by Alternative.me since February 2018, and it describes Bitcoin. Not "crypto". Bitcoin.
That single fact already clears up half of the confusion around it.
How the index is built
Alternative.me publishes the recipe. Every component is compared against its own 30- and 90-day averages, so the index reacts to unusual behaviour rather than absolute levels.
| Component | Weight | What it looks at |
|---|---|---|
| Volatility | 25% | Current volatility and max drawdown vs. 30/90-day averages. A spike counts as fear. |
| Momentum & volume | 25% | Buying volume in a rising market vs. the same averages. High volume in green markets counts as greed. |
| Social media | 15% | Post counts and interaction speed on Bitcoin hashtags. Unusually high engagement counts as greed. |
| Surveys | 15% | Weekly polls with 2,000 to 3,000 votes. Currently paused. |
| Bitcoin dominance | 10% | Rising BTC share of total market cap reads as a flight to safety (fear); falling dominance as risk appetite (greed). |
| Google Trends | 10% | Search volume changes and related queries. "bitcoin price manipulation" trending is fear; "how to buy bitcoin" is greed. |
The result lands on a 0 to 100 scale. Our page labels it like this: 0 to 24 is Extreme Fear, 25 to 44 Fear, 45 to 55 Neutral, 56 to 74 Greed, and 75 or higher Extreme Greed. Other sites draw the lines a few points differently. Those cut-offs are conventions, not thresholds anyone tested for significance.
Two things worth noticing in the table. Half the index is derived directly from price and volume data, so it partly measures what the chart already shows. And with surveys paused, 15% of the published weighting is not in the live number; how that weight is redistributed is not documented.
What the extremes meant historically
The index has printed 10 or lower on 69 of the 3,143 daily readings since February 2018, spread across 17 separate episodes (a gap of more than a week counts as a new one). The older episodes: around the December 2018 low near $3,200, during the March 2020 crash (down to 8), and through the 2022 deleveraging, with 8 in May after the Terra collapse and 6 in June during the Celsius unwind. The recent ones matter just as much. 32 of those 69 days fall after January 2025, nearly all of them in 2026: between early February and early March 2026, 19 daily readings landed at or below 10, spread across a five-week window rather than back to back, among them the all-time low of 5, a level previously seen only in August 2019; a smaller cluster of seven such days followed from late March into early April, and three more in June 2026 (as of mid-September 2026). Extreme fear is not a museum piece. In hindsight, most of those episodes clustered around a multi-month or cycle low. Not all of them: the FTX collapse in November 2022 never pushed the index below 20, even though the price set its cycle low that month.
The greedy end is less tidy. The index hit 95 in February 2021 with Bitcoin near $58,000. The price kept climbing for two more months before the April top. When Bitcoin set its then all-time high near $69,000 in November 2021, the index peaked at 84, high but noticeably below the 95 it had printed nine months earlier at a lower price. Sentiment was already cooling while price made new highs.
Our Fear & Greed page also carries a small table with rounded reference values per zone: roughly how Bitcoin moved over the following 30 and 90 days in past episodes. Read it as orders of magnitude, not as a backtest. The figures are rounded, cover a small number of episodes, come without a published sample size, and the spread around them is enormous. The page says as much in its own footnote.
Four ways people misread it
1. Treating it as a timing signal
"Extreme fear, time to act" is the most common misuse. The index sat at or below 20 for weeks in late 2018 and for 62 consecutive days in 2022, from early May to early July. Anyone who reacted to the first sub-20 print that May watched Bitcoin lose close to half its value over the following six weeks while the index stayed pinned. It describes a regime, not a day.
2. Applying it to altcoins
The volatility, momentum and dominance inputs are Bitcoin-specific. An altcoin can drop 60% while the index reads Neutral, because Bitcoin did nothing that week. If you hold anything other than BTC, the index tells you about the tide, not about your boat.
3. Reading it as leading
Because half the weighting comes from price and volume, the index tends to confirm what already happened. A crash produces volatility, volatility produces a low reading. That is not a forecast; it is a summary of the last few days with a sentiment label attached.
4. Assuming 50 means "balanced"
Neutral is a band, not a fair-value estimate. A reading of 48 after three weeks of 80s means something very different from 48 after three weeks of 15s. Direction and duration carry more information than the level. That is why our page puts a history chart right under the current gauge, switchable from 30 days up to the full dataset. Look at the line before you look at the needle.
Combining it with other signals
On its own the index is a mood thermometer. It gets useful when you line it up with data that measures something else.
- Bitcoin dominance and total market cap on our market overview. Fear with rising dominance is a different regime from fear with falling dominance (the latter usually means altcoins are being sold harder than BTC).
- Volume. A low reading on thin volume is apathy. A low reading on record volume is capitulation. The index does not distinguish between the two; the volume figure does.
- Macro context. The FRED indicators on the market overview (Fed funds rate, CPI, dollar index) explain a lot of the multi-month fear regimes since 2022. Crypto sentiment rarely turns while those are moving against it.
- Duration in zone. Count the days. Three days of Extreme Fear after a liquidation cascade is noise. Six weeks of it is a regime.
A simple way to use it: note the zone, note the direction over the last two weeks, note whether volume and dominance agree. If all three point the same way, the label is probably describing something real. If they disagree, the label is the weakest of the three.
What the index is good for
It is a fast, consistent, free reference for "how does the crowd feel about Bitcoin right now", with eight years of daily history behind it. That is genuinely useful for putting your own reaction in context. Feeling anxious when the index reads 12 is the crowd's state, not information about the future.
What it is not: a signal generator, an altcoin indicator, or a substitute for looking at the actual market data. Alternative.me says as much in its own disclaimer.
Check the current reading and the full history on our Fear & Greed page. The data updates daily from Alternative.me.
This article is for information purposes only and is not financial advice. Historical patterns describe the past and do not predict future market behaviour.