ICP: CryptoKeyhole Fear & Greed Index

Internet Computer CryptoKeyhole Fear & Greed Index was 47.69 on 9 Nov 2026 (latest daily value).

A 0-100 reading of fear and greed in the Internet Computer market, from its own price, derivatives and attention data, each part judged against its own past two years.

Why use it. Markets swing between fear, when holders sell at any price, and greed, when buyers chase. Extremes in either direction have tended to come near turning points, so the reading helps put a move in context. How to read it. Below 25 is extreme fear, 25-44 fear, 45-55 neutral, 56-75 greed and above 75 extreme greed. Each component is ranked against its own past two years, so 80 means today's mix is greedier than about four days in five of that period. The component chart shows which part is driving the reading. It can read differently from other fear and greed indexes: derivatives positioning (funding, open interest, taker flow and the long/short ratio) carries 40 of its 100 weight points and every part is ranked against its own past two years, while widely quoted market-wide indexes lean on price momentum, volatility and social or survey sentiment. When traders add leverage and pay to stay long into a falling price, this index can read greedier than those; the component chart shows that split.

Calculation. A 0-100 sentiment reading for Internet Computer from 5 components of CryptoKeyhole's own data: price momentum, half (close / 30-day average - 1) plus half (close / 90-day average - 1), from the coin's own daily USD price closes; volatility, the standard deviation of the last 30 daily log returns, annualised; inverted, since unusually high volatility comes with fear; funding rates, the 7-day mean of the daily median across venues of each venue's perpetual funding rate (8-hour rate), from our Futures Perpetual Funding Rate (All) chart; taker flow, the share of Binance spot and perpetual volume bought by takers over 7 days (the mean of the two markets); the long/short account ratio, each venue's 7-day mean ranked against that venue's own past two years and the venues' scores averaged, so a venue whose collection began recently joins once it has a year of readings. Each component is scaled to 0-100 by its own history: the percentile rank of the day's raw reading among the coin's raw readings of the trailing 730 days, the day itself included, needing at least 365 of them: 100 x (readings below + half the other equal readings) / (readings - 1). Windows need most of their days (27 of 30, 81 of 90, 5 of 7). The index is the weighted mean of the components that score that day, with their weights rescaled to sum to 100; it is published when price momentum and at least two more components score and they carry at least half of the weight. Weights: momentum 25, volatility 15, funding 15, open interest 10, taker flow 10, long/short ratio 5, holder profit 10, exchange flows 5, attention 5, options implied volatility 10 (Bitcoin and Ether). No look-ahead: every window ends on the day it scores, and a late input (Wikipedia views) carries its latest score for at most three days. History is recalculated from the charts as they are published now, so a later correction of an input also corrects the days it touches. The open UTC day is a partial from each input's day so far, labelled partial until the day ends. Zones: 0-24 extreme fear, 25-44 fear, 45-55 neutral, 56-75 greed, 76-100 extreme greed. The hourly view applies the same method to completed UTC hours: momentum from the 30- and 90-hour averages, volatility from the last 30 hourly returns, and funding, open interest, taker flow, the long/short ratio and options implied volatility over hours, each ranked against its own trailing 730 hours (at least 365) and carried for at most three hours when late. Parts with daily inputs only (holder profit, exchange flows, attention) are left out of it and the weights rescale; each hour appears once it has ended and its price is published. Not available for ICP: open interest, holder profit, exchange flows, attention, options implied volatility (no input with enough history), so their weight goes to the others.