How Is the Fear and Greed Index Calculated?

The Basic Method

The calculation has three steps. First, measure each signal, how volatile the market is, how strong momentum is, how active social chatter is, and so on. Second, score each signal on its own fear-to-greed scale. Third, combine them, usually as a weighted blend, into one number from 0 to 100. The point of using many signals is robustness: a Fear and Greed Index that leaned on one input would just track that input. Blending several means the score reflects the overall mood, not a single metric.

The Three Steps In Detail

Each step does real work. Measuring means pulling raw, live data for every input, prices, trading volume, volatility, social posts and more. Scoring is the clever part: each raw number is "normalised", compared against its own recent history, so it becomes a fear-to-greed reading rather than an absolute figure. This matters because a volatility level that is high for one asset may be normal for another; normalising lets very different signals speak the same language. Finally, the normalised signals are blended into a single 0 to 100 score. Because each input is judged relative to its own baseline, the index measures how unusual today’s mood is, not just the raw numbers.

The 10 Indicators CFGI Reads

CFGI combines ten distinct families of signal, each capturing a different face of the crowd’s emotion.

• Price momentum: how far price sits above or below its recent averages.
• [Volatility](/learn/markets/what-is-volatility/): rising volatility signals fear, calm signals confidence.
• [Trading](/learn/markets/what-is-trading/) volume: how much activity is behind the move.
• Social activity: the volume and tone of chatter about an asset.
• Market dominance: shifts between Bitcoin and the rest of the market.
• Search interest: how intensely the public is searching for it.
• On-chain signals: whale movements, order-book pressure and similar.

Each is scored and weighted into the final reading, recomputed every 15 minutes for crypto and daily for stocks.

Why Blend So Many Signals

The blending is the whole point, for two reasons. First, accuracy: any single indicator is partial and noisy, social chatter can be gamed, a price move alone says nothing about conviction, so combining many produces a steadier, fuller picture than any one could. Second, resistance to manipulation: it is far harder to distort a ten-input gauge than a one-input one, because you would have to move many independent signals at once. This is why a synthesised reading at an extreme carries weight: reaching it required many separate measures of fear, or greed, to agree, not a single dramatic data point.

Per Asset, Per Timeframe

CFGI does not stop at one market-wide figure. It runs the whole calculation separately for each of more than 100 crypto assets and for stocks, and across four timeframes from short-term to long-term. That granularity is a deliberate design choice: it lets the index show that Bitcoin can be greedy while another coin is fearful, or that the short-term mood differs from the long-term one. A classic single-number index averages all of that away; computing per asset and per timeframe preserves exactly the differences that carry the most information.

What the Calculation Does and Does Not Do

It is worth being clear about what all this measuring produces. The calculation is designed to capture the present state of sentiment accurately, how fearful or greedy the crowd is right now, drawn from real, current behaviour. What it is not is a forecast. No amount of blending turns a measurement of today’s mood into a reliable prediction of tomorrow’s price; the index reads the present, and sentiment can stay extreme for a long time. So the calculation should be judged on the right terms: it is a careful, manipulation-resistant snapshot of crowd emotion, not a crystal ball. That distinction, between accurately describing now and predicting next, is the single most important thing to understand about how the number is built and how it should be used.