Fear and Greed Indexes: What the Scores Mean

Table of Contents

    A fear and greed index turns several market signals into one sentiment score. A low score tells you the inputs look cautious by that provider’s rules. It doesn’t tell you that prices are cheap or that a rebound is due. Before you act on a headline reading, identify the market it covers and look at which inputs moved. A drop driven by more expensive options protection raises a different question from a drop accompanied by weakness across stocks and corporate bonds. [1] [6]

    Start with the provider name. CNN’s stock-market gauge, Alternative.me’s Bitcoin gauge, and CoinMarketCap’s broader crypto gauge all describe sentiment on a 0–100 scale. A 20 on one index and a 20 on another don’t measure the same conditions, even if both providers use the word “fear.” [1] [3] [4]

    Stock and crypto fear and greed indexes

    Check the label, the assets covered, and the reading’s timestamp. Even two services reporting their own gauges on the same calendar day may be summarizing different markets or observation periods. The methods below are described as of the dates shown in the sources; these are not current readings.

    What each fear and greed index measures
    Provider Market Examples of inputs
    CNN Business U.S. stock-market risk appetite CNN’s reporting describes seven signals, including the VIX, stock-market momentum, options activity, and demand for safer assets and high-yield bonds. [1] [2]
    Alternative.me Bitcoin sentiment Bitcoin volatility and drawdowns, momentum and volume, social activity, Bitcoin’s share of crypto market value, and search trends. The provider lists surveys separately and marks them paused. [3]
    CoinMarketCap Wider crypto market sentiment Momentum among the ten largest cryptocurrencies excluding stablecoins, Bitcoin and Ethereum options measures, Bitcoin’s market value relative to major stablecoins, and CoinMarketCap’s own search and engagement data. [4]

    What the provider tells you about the calculation

    CNN’s April 2025 explanation and its November 2021 coverage describe several ingredients of the U.S. gauge. They don’t provide enough detail to reproduce the calculation, so the example below uses a separately defined method. An input can come from another asset class: demand for Treasuries or high-yield bonds can help describe risk appetite around stocks without making this a bond sentiment index. [1] [2]

    Alternative.me’s methodology, checked November 17, 2026, describes its published index as Bitcoin-only despite the broader “crypto” label. It compares its volatility and momentum measures with 30- and 90-day history. Its page also says Reddit sentiment analysis is outside the live index and labels its nominal 15% survey input paused. The page doesn’t explain how that 15% is handled in the published reading. Dropping the input and scaling the other weights, or filling the gap with a neutral value, would be two different assumptions. CoinMarketCap’s published component descriptions also leave the exact blend unspecified. [3] [4]

    How a fear and greed index is built

    Market breadth, implied volatility, and credit spreads use different units. Breadth measures how widely stocks participate in a move; one way to track it is the percentage of stocks above their own moving averages. The VIX reflects near-term volatility implied by S&P 500 options. A high-yield credit spread measures the extra yield on lower-rated bonds relative to a benchmark, with the precise definition depending on the series. Averaging a percentage of stocks, a VIX reading and a yield gap as raw numbers would give their units an arbitrary influence on the result. [6] [8]

    1. Choose the measures and their reference periods. Define the assets, data sources and observation cutoff. Decide whether each input is compared with a trailing window or another baseline. Comparing against recent history answers a different question from comparing against a long fixed history.
    2. Put inputs on a common scale and align their direction. One approach is to rank each value against its own history. If high breadth is treated as a sign of confidence while high implied volatility is treated as a sign of caution, reverse the volatility ranking before combining the scores.
    3. Combine the scores and define how to handle gaps. A simple weighted average is one choice. Document the weights and what happens if a value is stale or missing. Reweighting changes the mix, and substituting a neutral score treats an unknown value like an observed middle reading.

    Worked example: from three inputs to one score

    Suppose you check the stock market after the daily inputs are available. You have the current breadth reading, the VIX level and a high-yield credit spread, plus 20 prior observations for each taken at comparable cutoffs. Breadth here is the share of stocks in your chosen universe trading above their own 50-day moving averages. All values are hypothetical, none are tied with the current reading, and the three signals count equally.

    Illustrative 0–100 component score
    Score = 100 × (prior observations more fear-like than today ÷ 20)
    Lower breadth, a higher VIX level, and a wider credit spread count as more fear-like in this example.
    One hypothetical market snapshot
    Input Current value More fear-like observations Score
    Breadth 38% 4 of 20 were lower 20
    VIX 28 points 2 of 20 were higher 10
    High-yield credit spread 4.5 percentage points 7 of 20 were wider 35

    The composite is (20 + 10 + 35) ÷ 3 = 21.7, rounded to one decimal. Its most cautious input, compared with its own recent history, is implied volatility. Credit looks less stressed by the same ranking rule. If breadth alone improved to a component score of 50 and the other scores stayed fixed, the average would rise to (50 + 10 + 35) ÷ 3 = 31.7. A ten-point gain in the headline would then come from just one input. The example does not reproduce any provider’s index or assign either result to a provider’s “fear” category.

    How to interpret an extreme reading

    On the afternoon of April 4, 2025, CNN reported a reading of 4 during a broad selloff. In November 2021, CNN reported extreme greed alongside record highs in major U.S. stock indexes. In each case, the gauge described the conditions investors were facing at the time. Those observations alone tell you nothing about what a strategy based on the label would have earned. [1] [2]

    A contrarian reading asks whether prices have moved further than new evidence warrants. If a gauge falls while the earnings outlook worsens and borrowing costs rise, there may be good reasons for lower prices. If a few headline indexes hold up but fewer stocks are participating, the apparently calm headline can also hide weakness. A high score can reflect improving business conditions or overconfidence. Examine the evidence for each interpretation before attributing the reading to herding or other investor biases.

    1. Check what changed. Read the component breakdown where available. Compare equity indexes, participation across sectors, and corporate credit on a consistent time basis. If the provider doesn’t show the components, don’t claim to know what drove its reading.
    2. Check what the score leaves out. For equities, compare prices with earnings expectations and recent changes to them. For a Bitcoin gauge, don’t substitute news about the S&P 500 or assume the result represents every token.
    3. Decide what would change your view. A broadening decline, a change in the earnings outlook, or a further move in credit may matter more to your analysis than a change in the gauge’s label. Use the score to choose what to investigate.

    How indexes differ from sentiment surveys

    The AAII Investor Sentiment Survey asks participants whether they think the stock market will rise, fall or stay roughly unchanged over the next six months. Its weekly bullish, bearish and neutral readings are shares of respondents. The members who choose to answer are a particular sample. A gauge built from market inputs, by contrast, infers sentiment from what prices, options and other measures are doing. An index value of 20 isn’t equivalent to a survey in which 20% of respondents are bullish. [5]

    Even a direct survey of expectations has limited value for timing. AAII reports that just 23.9% of respondents were bullish in its July 3, 2008 observation, an unusually low share. The S&P 500 then fell 28.4% over 26 weeks. That is one counterexample to an automatic contrarian rule, not a test of the CNN or crypto indexes. If a survey and a market gauge disagree, check the survey’s question and sample and the composite’s inputs before trying to reconcile them. They answer different questions. [5]

    Limitations and common mistakes

    • Counting several reactions as independent evidence. A single shock can pull stock prices down, lift implied volatility and widen credit spreads. A composite can describe the breadth of that response without supplying three independent reasons to expect a reversal.
    • Treating the label as a valuation or probability. “Extreme fear” describes a reading under the provider’s rules. It doesn’t say how large a selloff is left or how likely a gain is. Likewise, a high VIX is about expected movement, not whether the S&P 500 will rise or fall. [6]
    • Splicing series that use different rules. Do not fill a missing CNN date with a crypto score or treat a revised history as what investors saw in real time. Check whether the inputs, methodology or update schedule changed before comparing extremes. [3] [4] [5]

    What would a fair timing test require?

    Choose the provider, threshold, investment, entry timing and holding period before judging results. Use the score that was actually available before each assumed trade. For example, a close-based reading posted after the close cannot support an entry at that same close. If you choose the threshold using earlier data, freeze the rule before testing it on a later period you haven’t inspected. Trying dozens of thresholds and reporting the best one can make chance look like skill. [5] [7]

    Compare total return and peak-to-trough losses against a relevant benchmark over the same period. For an equity timing rule, a buy-and-hold comparison should include dividends; the timing rule should also account for time in cash and trading costs. Ten consecutive fear readings during one crisis don’t give you ten independent episodes, especially if their forward return windows overlap. A sensible test checks whether results depend on one event or market regime. The same discipline applies when testing a separate momentum investing strategy. [7]

    Frequently Asked Questions

    It is the numerical midpoint of a 0–100 scale. Read the provider’s category label and components: the midpoint can reflect mixed signals or several middling inputs. It isn’t a 50% chance of a gain.

    Alternative.me says its published index is for Bitcoin. CoinMarketCap combines a broader momentum input with Bitcoin, Ethereum and other data. Check the provider, since the label alone doesn’t define the universe. [3] [4]

    The VIX is a measure of near-term volatility implied by S&P 500 option prices. CNN describes it as one of several inputs to its sentiment reading. The two can tell different stories because the composite has other inputs. [1] [6]

    The label describes how the provider’s inputs line up. It can prompt you to investigate what changed in prices, earnings expectations or credit, but it doesn’t establish that the market is underpriced or that a rebound will come within your holding period.

    Sources

    1. Allison Morrow, CNN: A brief history of CNN’s Fear and Greed Index, April 7, 2025, syndicated by CNN Newsource.
    2. Paul R. La Monica, CNN Business: There is no fear on Wall Street. Only greed, November 9, 2021, syndicated by CNN Newsource.
    3. Alternative.me: Crypto Fear & Greed Index methodology, accessed November 17, 2026.
    4. CoinMarketCap: CMC Crypto Fear and Greed Index methodology, accessed November 17, 2026.
    5. Charles Rotblut, AAII: Analyzing the AAII Sentiment Survey Without Hindsight, 2014.
    6. Cboe: Volatility Index overview, accessed November 17, 2026.
    7. Bailey, Borwein, López de Prado and Zhu: The Probability of Backtest Overfitting, February 2015 revision.
    8. ICE Data Indices via FRED: ICE BofA US High Yield Index Option-Adjusted Spread, series notes, accessed November 17, 2026.
    Disclaimer

    This article is for education and does not provide individual investment advice. The worked composite is hypothetical. The historical cases illustrate how to interpret a reading and do not establish a trading rule.