Fear & Greed Hits 71: The Black-Box Math Behind Crypto's Most-Quoted Number

71.

That's the number that hit my terminal at the last daily close. The Crypto Fear & Greed Index — the composite sentiment gauge retail traders treat like a weather forecast — printed 71, up 7 points in a single day.

Greed. Not extreme greed. Not yet. Just... greed. The kind that feels reasonable. The kind that lets you tell yourself you're being disciplined when you're really just early to your own exit.

Within two hours, my feed was flooded. "Bullish confirmation." "Risk-on is back." "We're so back." The same accounts posting red candle doom memes last month were suddenly charting the road to a new all-time high. One thread pulled 40,000 likes before I finished my coffee.

Here's what nobody posted: the 7-day average sits at 70. The 30-day average sits at 67. And the current reading — 71 — is only 4 points from "extreme greed."

I've done 7x24 market surveillance long enough to recognize a monotonic climb. And I've watched enough of them end badly to know the most dangerous number in crypto isn't 100. It's the number right before the crowd decides it can't go higher.

Let me show you what this number actually is — and what it isn't.

Context: How a Single Integer Gets Built

First, the basics, because half the people quoting this index couldn't tell you how it's constructed.

The Fear & Greed Index is a 0-to-100 composite. Lower means fear, higher means greed. The bands are standardized across most implementations: 0-24 extreme fear, 25-49 fear, 50-74 greed, 75-100 extreme greed. This reading came from Coinglass, the derivatives data platform that's become the default tape for anyone trading perps.

The classic model — the one that made the index famous — comes from alternative.me. That version bundles six inputs: volatility at 25%, market momentum and volume at 25%, social media sentiment at 15%, surveys at 15%, Bitcoin dominance at 10%, and Google Trends at 10%. Weighted, blended, smoothed, and spat out as a single integer.

Coinglass runs its own version. It aggregates exchange price action, volume, and social signals into a comparable composite.

Notice what I just said. Comparable. Not identical.

That distinction matters more than the headline. Two platforms can look at the same market on the same day and print different readings, because they weight the inputs differently. The F&G index isn't a measurement. It's an opinion dressed up as a measurement.

And here's the thing about this snapshot. It arrived with a payload of exactly five numbers: 71, plus 7, seven-day average 70, thirty-day average 67, and the label "greed."

That's it. No funding rates. No open interest. No stablecoin flows. No on-chain transfer data. No price context.

Just a sentiment integer and its moving averages.

I've spent years translating technical documents for retail readers, and the most important skill isn't reading the number. It's knowing which numbers are missing. So let's start there.

Core: What the Data Actually Says

The monotonic climb is the story — not the level

Everyone's fixated on the 71. The greed label. But the more interesting signal is the shape of the curve.

Current: 71. Seven-day: 70. Thirty-day: 67.

Read that top to bottom and you get a clean staircase. 67 → 70 → 71. Each window higher than the last. That's not a spike. That's not a one-day emotion dump. That's a sustained, multi-week warming trend.

I've modeled enough sentiment cycles to know the difference. A single-day jump tells you about an event — a headline, a liquidation cascade, a surprise ETF print. A monotonic climb across three timeframes tells you about a regime. The crowd has been getting progressively more comfortable with risk for weeks.

Which cuts two ways.

Bullish reading: the trend is your friend. Sentiment is improving, and improving sentiment attracts marginal capital. Simple.

Bearish reading: sentiment is a lagging indicator, and a lagging indicator climbing for weeks is a lagging indicator confirming a move that already happened. You're not early. You're late.

The +7 jump is a fact about the past, not the future

Let me be precise, because this is where 90% of the misinterpretation lives.

The F&G index is built from realized data. Prices that already printed. Volumes that already traded. Posts that already went up. By construction, it describes what happened. It does not predict what happens next.

A +7 daily move tells you the market got materially more euphoric in the last 24 hours. It does not tell you whether that euphoria continues, reverses, or chops sideways for a month.

I learned this the hard way during the 2020 DeFi summer. I was running weekly Spaces on yield farming risk, and I built a real-time model of impermanent loss across the Curve pools. I watched sentiment metrics scream "greed" for three straight weeks before a major protocol exploit drained the party. The index didn't warn me. The on-chain flows did — and only because I was staring at them instead of the sentiment integer.

The F&G index is a rearview mirror with a nice paint job.

The methodology is a black box, and that's a structural risk

Here's the part that should make you uncomfortable.

The classic alternative.me model publishes its factor weights. Volatility 25%. Momentum 25%. Social 15%. Surveys 15%. Dominance 10%. Trends 10%. You can audit it. You can argue with it. You can rebuild it.

The Coinglass weights? Not disclosed in this snapshot.

I'm not accusing anyone of manipulation. That's not the point. The point is that an undisclosed weighting scheme is an unauditable one. And an unauditable sentiment index is a number you take on faith.

I've audited enough protocols to have a reflex about this. When I infiltrated those ICO Telegram groups back in 2017, the tell was always identical: beautiful whitepapers, zero GitHub commits. The absence of verifiable detail wasn't a coincidence. It was the strategy.

Sentiment indices aren't ICOs. But the principle holds. If you can't see how the sausage gets weighted, you're not analyzing the index. You're trusting the butcher.

The missing data is the actual headline

Let me list what normally accompanies a sentiment reading like this — and what's conspicuously absent.

Funding rates. In a healthy greed regime, perp funding goes modestly positive. In an overheated one, it spikes. I've seen funding hit annualized triple digits right before a cascade. Without it, I can't tell whether this greed is leveraged or spot-driven.

Open interest. OI tells you how much money sits on the table. Rising OI into rising greed is a recipe for a violent unwind. Falling OI into rising greed is a healthier picture. We have neither number.

Stablecoin net flows. The cleanest proxy for fresh capital. Greed with inflows is accumulation. Greed without inflows is rotation — money moving around inside a closed casino. Critical distinction. Absent.

Large on-chain transfers. Whale behavior. When the index hits greed and whales move coins to exchanges, that's distribution. When they move to cold storage, that's conviction. Unknown here.

I spent a chunk of early 2022 doing exactly this — tracking wallet movements during the NFT floor crash, identifying the specific addresses dumping into retail bids. The lesson wasn't "NFTs are risky." The lesson was that sentiment is the cover story and flows are the actual story. When a floor drops 40% in a day, it's never the vibe that moved first. It's always the wallets.

Right now, on this data, I can't see the wallets. I can only see the vibe.

The self-reinforcing loop nobody prices in

Here's a second-order effect I rarely see discussed, and it's the one that keeps me up during surveillance shifts.

Sentiment indices don't just measure the market. They shape it.

When a widely-followed account posts "Greed index hits 71 — bullish," that post is itself a sentiment input. Retail reads it, feels validated, buys. Volume ticks up. Social engagement ticks up. And those are exactly the inputs the index weights. The number nudges the market that produces the number.

This is the feedback loop. In thin, low-liquidity conditions — which is most of crypto's weekend tape — it can run faster than any fundamental driver.

Wash trading is the digital casino's tell, and the house doesn't even need to rig the game. The players do it to themselves.

Now, the counterargument. The classic rejoinder is that this is just reflexivity, and reflexivity is a feature, not a bug. George Soros made a career out of it. Fair. But Soros also knew when to exit a reflexive trend, and the signal he used was never "the sentiment gauge says I'm right."

Stress-testing the "extreme greed is close" narrative

The headline framing you'll see is: "71, only 4 points from extreme greed, buckle up."

Two things about that.

First, thresholds are arbitrary. The 75 line isn't physics. It's a convention — a round number someone picked because it looked clean on a chart. Treating 75 as a magic door is like treating a stock at $99.99 as fundamentally different from one at $100.01. The market doesn't know about your thresholds.

Second, and more importantly: the gap between "greed" and "extreme greed" is where the most dangerous psychology lives. It's the zone where people tell themselves they're being disciplined — "I'm not greedy, I'm just positioned" — while acting exactly like the top-tick buyers they mock.

I've watched this psychology in NFT communities. The floor is down 20% and people say "healthy correction." Down 30% — "accumulation zone." Down 40% — silence, then the wallet dumps. The labels change. The behavior doesn't.

Historical base rates — handle with care

Every contrarian piece cites "historically, extreme greed precedes pullbacks." True in a loose, unfalsifiable way. Also almost useless without base rates.

Here's the honest version. Sentiment indices are constructed from price and volume. Price and volume mean-revert. So of course high sentiment sometimes precedes pullbacks — you've basically built a momentum oscillator and then declared that momentum oscillators oscillate.

The real question isn't "does greed precede pullbacks." It's "conditional on greed at 71 with a rising trend, what's the distribution of forward returns?" That requires the funding, OI, and flow data we established is missing.

Anyone who tells you they know the answer from a single integer is selling you something.

The view from the surveillance seat

Let me give you the view from my seat, because it's different from the view on the timeline.

I work 7x24 surveillance. That means I don't see the market as daily candles. I see it as a stream of anomalies — order book imbalances, sudden volume clusters, wallets that move in patterns that don't look human.

When the sentiment index climbs monotonically, here's what I actually watch: wash trading. In a greed regime, volume inflates. Some is real. Some is the digital casino printing its own liquidity. When volume is up but order book depth is unchanged, that's a tell. The tape is being painted.

I've flagged wash patterns on enough exchanges to know sentiment indices can't see them. A wash trade registers as volume. Volume is an input. The index climbs on fake activity as happily as real activity. There's no lie detector in the composite.

So the 71 could be built, in part, on activity that doesn't represent genuine demand. I can't prove it from this data. But I can tell you it's not in the model's threat list.

And the exit liquidity question

Every greed reading has a hidden structure: someone is providing exit liquidity, and someone is taking it.

At 71 with a rising trend, the marginal buyer is the late buyer. The person who read the bullish thread. The person who's been waiting for "confirmation" and just got it. Confirmation is what tops are made of — because by the time a signal is loud enough for the cautious crowd, the informed crowd has already positioned.

Exit liquidity is someone else. Always. The only question is which side of the trade you're on when you find out.

One source, one point of failure

The snapshot cites Coinglass. Single source. That's a supply-chain risk in itself.

If Coinglass adjusts its weighting — say, to cut social media weight after a manipulation scandal, or to add a new factor — the index re-bases. Every downstream algorithm, every trading bot using this index as a filter, every newsletter quoting it, shifts overnight. Not because the market moved. Because the ruler changed length.

This is the indicator drift problem, and it's chronic in crypto data. I've cross-checked alternative.me against Coinglass on the same day and seen readings diverge by 10+ points. Same market. Same timestamp. Different answer. Because they're different instruments wearing the same name.

If you're using sentiment as a signal, know which instrument you're reading — and check more than one.

Contrarian: It's Not a Thermometer, It's a Mirror

Here's the angle you won't find in the threads.

Everyone treats the F&G index as a thermometer. It's not. It's a mirror.

A thermometer tells you the temperature. A mirror tells you what the crowd looks like. And the most important thing a mirror reveals isn't the crowd's mood — it's who's holding the mirror.

The F&G index at 71 doesn't tell you the market is greedy. It tells you the data feeding the index was produced in a greedy environment. Those are different claims. The first is a fact about the world. The second is a fact about the data pipeline.

And in a bear market — which is where we are, structurally, regardless of what a sentiment integer says — the gap between those two claims is where the traps live.

Think about it. We're in a market that has spent months grinding lower. Now the sentiment gauge climbs toward extreme greed. Two readings:

Reading one: the bear is over, sentiment is leading the turn. Bullish.

Reading two: this is a bear market rally, and the sentiment index is doing exactly what it's designed to do — telling you the crowd is getting excited right before it gets hurt. Bearish.

The index cannot distinguish between these. It's structurally incapable. Because it doesn't know about the bear market. It only knows prices, volumes, and posts. And in a bear rally, all three look great — right up until they don't.

I've seen this dynamic before. In early 2022, during the NFT crash, sentiment spiked on every bounce. Greed readings printed on dead-cat rallies. Each one was a fresh opportunity for whales to distribute into retail optimism. The wallets knew. The sentiment gauge didn't.

Here's the contrarian conclusion, and I want you to sit with it:

The more confident the sentiment gauge gets, the less it should affect your position sizing — and the more it should affect your risk controls.

That's backwards from how almost everyone uses it. Retail treats a rising greed index as a green light to add. The surveillance view treats a rising greed index as a yellow light to tighten stops, cut leverage, and watch funding rates like a hawk.

The crowd sees the number and feels permission. I see the number and feel the exits narrowing.

And the final twist. The people amplifying this number loudest are the ones with the most to gain from you acting on it. Media accounts need engagement. Influencers need conviction validated. Data platforms need their indices quoted. None are lying, exactly. They're all pointing at the same mirror and calling it a window.

Wash trading: The digital casino's tell. Narrative amplification is its marketing department. And the greed index at 71 is the flyer they hand you at the door.

Takeaway: Watch the Gap, Not the Level

So what do I actually do with a 71?

I don't trade it. I track it. Specifically, I track what's missing around it.

The number that matters isn't 71. It's the funding rate that hasn't been published. The open interest nobody's quoting. The stablecoin flows that would tell me whether this greed is funded or borrowed.

If the index pushes through 75 into extreme greed while price stalls, that's a divergence — and divergences are where the real information lives. If it pushes through 75 and price confirms with fresh inflows and modest funding, that's a genuine risk-on regime.

Same number. Opposite conclusions. The difference lives entirely in the data this snapshot doesn't have.

My forward-looking judgment: watch the gap, not the level. Watch whether the sentiment index and the hard data — funding, OI, flows — agree or diverge. When they agree, the trend is real. When they diverge, the trend is a story, and stories end.

Red candles don't announce themselves. They arrive on the most bullish day of the month, when the greed index is highest and the crowd is loudest.

71 isn't a warning. It's an invitation to look closer.

And the exit liquidity? It's someone else. Make sure it isn't you.