Ask a crypto trader how the market "feels" and you will get an opinion. Ask the Crypto Fear & Greed Index and you get a number between 0 and 100. That simplicity explains why the gauge is quoted constantly in market commentary — and why it is so frequently misread. The index is not a forecast, it is not a valuation model, and it does not measure the whole crypto market. Understanding what actually goes into the score is the difference between using it as a useful check on your own psychology and mistaking it for a trading signal.
What the index actually measures
The best-known version is published by Alternative.me, which launched it in 2018 after Bitcoin's volatile 2017 run, taking inspiration from CNN Money's long-running Fear & Greed Index for equities. It refreshes once a day at 00:00 UTC, so intraday swings do not move it.
The published methodology combines the following inputs, each compared against its own recent averages rather than judged in isolation:
- Volatility (25%) — current Bitcoin volatility and maximum drawdowns measured against 30-day and 90-day averages. Unusual volatility is treated as a sign of a fearful market.
- Market momentum and volume (25%) — current buying volume and momentum against prior averages. Sustained heavy buying into a rising market pushes the score toward greed.
- Social media (15%) — interaction and engagement rates on Bitcoin-related posts. Unusually high engagement is read as rising greed.
- Bitcoin dominance (10%) — Bitcoin's share of total crypto market capitalisation, used as a proxy for risk appetite. Falling dominance implies speculative rotation into altcoins.
- Google Trends (10%) — search volume for Bitcoin-related queries, where the composition of terms matters as much as the volume.
The component that is no longer running
Here is a detail that rarely survives into secondhand explainers. Alternative.me's own page also lists a surveys component weighted at 15%, based on weekly public polls — and marks it as currently paused. It has not been feeding the live index.
The arithmetic matters. The five active inputs sum to 85%, not 100%. The provider does not publish how that remaining 15% is redistributed across the surviving components, and the underlying calculation is proprietary, meaning outside users cannot independently reproduce the daily score or verify the effective weightings. Anyone citing the index as a precisely engineered instrument is overstating what is publicly documented.
Reading the scale
The 0-100 range is conventionally divided into bands. As Forbes sets them out:
- 0-24 — extreme fear
- 25-49 — fear
- 50 — neutral
- 51-74 — greed
- 75-100 — extreme greed
Band boundaries vary slightly between publishers, so a score sitting near a threshold can be labelled differently depending on where you read it. The label is a presentation choice layered on top of the number; the number is what carries the information.
Why two indexes with the same name disagree
This is the most practical thing to understand, and it is easy to demonstrate. Checking both major providers on the same day in July 2026, Alternative.me showed 29, classified as Fear. CoinMarketCap's index showed 33, also in fear territory. Same concept, same scale, same day, different numbers.
The gap is not an error. CoinMarketCap built its index on a different set of inputs entirely:
- Price momentum across the top 10 cryptocurrencies by market cap, excluding stablecoins
- Volatility drawn from Volmex implied volatility indices for Bitcoin and Ethereum
- Derivatives positioning, via put/call ratios in options markets
- Market composition, using the Stablecoin Supply Ratio
- Proprietary platform data, including keyword searches and user engagement
Note how different these are in character. CoinMarketCap leans on forward-looking derivatives pricing — implied volatility and options skew reflect what traders are paying to hedge. Alternative.me leans on realised volatility, social chatter and search interest, which describe what has already happened. Two defensible designs, two different answers. Neither is the "true" reading, and quoting "the" Fear & Greed Index without naming the publisher is imprecise.
What the index cannot do
The index is backward-looking by construction. Realised volatility, past volume and search history all describe conditions that have already occurred. It characterises the present mood; it does not predict the next move.
It is also Bitcoin-centric. Volatility, dominance and search inputs are anchored to Bitcoin, so the score may say little about sentiment in altcoins, DeFi tokens or any specific asset you happen to hold.
The deepest problem is with the contrarian logic itself. The index rests on two assumptions: that extreme fear signals excessive worry and possible opportunity, and that extreme greed signals a market due for correction. Directionally reasonable, useless for timing. Markets can hold extreme readings for weeks at a stretch during a strong trend, and prices frequently keep moving in the same direction throughout. Extreme fear is not a bottom signal; extreme greed is not a top signal.
Backtesting work on the index has highlighted a related trap. Readings below 10 are genuinely rare — sometimes absent for years. A mechanical rule that sells into extreme greed and waits for extreme fear to re-enter can leave a trader parked in cash through long stretches of rising prices, waiting for a trigger that never fires. The index is best treated as one input alongside price structure, on-chain data and fundamentals, never as a standalone system.
Using it sensibly
The index's genuine value is behavioural rather than predictive. Its highest and best use is as a mirror. When the score sits at extreme greed and you find yourself increasing position sizes and shortening time horizons, the number is telling you something about your own state, not about tomorrow's price. When it sits in extreme fear and you are tempted to capitulate, the same applies in reverse.
Three habits make it more useful. Track the trend rather than fixating on the daily print, since the direction of travel over weeks carries more signal than a one-day move. Always name your source, because a reading of 29 and a reading of 33 came from different methodologies on the same afternoon. And pair it with hard data — funding rates, exchange flows, realised losses — before letting it influence a decision.
Used that way, a single number between 0 and 100 becomes what it was designed to be: a quick, imperfect read on whether the crowd is panicking or getting comfortable, and a prompt to check whether you are moving with it.
This article is for informational purposes only and does not constitute investment advice.