Fear and Greed, Used Without Over-Trusting It

The index reads 14. Extreme fear. Every article you open tells you extreme fear is where fortunes are made, and something in your stomach disagrees, and you would like to know which of the two is better informed.

The honest answer requires knowing what the number is built from, because most of the confidence attached to it comes from people who have never checked.

What actually goes into the number?

Alternative.me publishes the methodology, and it is worth reading rather than taking secondhand. Six inputs, with stated weightings:

Volatility carries 25%. It measures current Bitcoin volatility and drawdowns against 30 and 90 day averages, treating unusual volatility as a sign of fear. Market momentum and volume carries another 25%, comparing current buying volume and momentum against the same historical windows, with sustained high volume in a rising market read as greed.

Social media activity carries 15%, derived from interaction rates and hashtag engagement on Twitter. Surveys carry 15%, collected through weekly polls, and this component is currently paused, which means the live index is composed from the remaining five inputs rather than the six the documentation describes.

Bitcoin dominance carries 10%, on the reasoning that rising dominance reflects a flight to relative safety and falling dominance reflects appetite for speculation. Google Trends carries the last 10%, reading search volume for Bitcoin-related queries and the shape of what people are searching for.

The output is a single integer from 0 to 100, updated daily, calculated on Bitcoin alone. Alternative.me has flagged separate altcoin indices as forthcoming for some years.

What is the index measuring, and what is it not?

Here is the part that changes how you read it. Add up the inputs that are derived from market data rather than from people saying things: volatility at 25%, momentum and volume at 25%, dominance at 10%. That is 60% of the stated methodology, and with surveys paused it is a larger share of the live number.

Those inputs are computed from the same price and volume series you can see on a chart. The index is not independently measuring how investors feel. For the majority of its weight it is restating recent price action in the vocabulary of emotion, and then handing the restatement back to you as though it were a separate observation.

This is why the index and the chart so rarely disagree. It is not confirmation. It is the same data twice.

The genuinely sentiment-derived portion is small: social media engagement and search interest, together 25% of the stated weights. Both are worth something and both have known problems. Twitter engagement measures a self-selected population that skews heavily toward the loudest holders, and it is directly purchasable. Google Trends measures newcomer curiosity, which historically rises after a move rather than before one, which makes it a lagging input by construction.

None of this makes the index worthless. It makes it a smoothed, normalised summary of recent price behaviour with a sentiment garnish, which is a legitimate thing to be and a very different thing from what most people think they are reading. The comparison worth drawing is with an oscillator like RSI, which is also a normalised transformation of price and which nobody mistakes for a poll.

Where does it fail as a timing tool?

In the way every mean-reversion indicator fails, and one way that is specific to it.

An extreme reading tells you a condition is unusual relative to the recent past. It contains no information about how long the condition can persist or how far it can extend. During prolonged drawdowns the index has sat in extreme fear for weeks on end while price fell considerably further, and anyone who treated the first extreme fear print as a signal spent the rest of the decline being early. The same asymmetry runs the other way. Extreme greed appears early in strong uptrends and can stay there through the most productive part of the move, so selling the first greed reading systematically gives up the trend.

The measurement window causes a subtler failure. Because volatility and momentum are scored against 30 and 90 day averages, the index measures deviation from the recent past rather than absolute condition. A market that has been falling steadily for two months recalibrates: the fall becomes the new normal, and the index drifts back toward neutral while the drawdown continues. The number is quietly telling you "this is now typical," and it reads as "things have improved."

Two more limits deserve stating. The index is Bitcoin. If your holdings are mostly altcoins, you are reading a proxy that decouples exactly when it matters, and Bitcoin dominance is a better tool for the rotation question it is being used to answer. And the index has become popular enough that its readings are themselves an input to behaviour, which erodes whatever edge a widely watched contrarian gauge ever had.

What is it genuinely good for?

Two things, both of which are about you rather than about the market.

It is a calibration check on your own state. If the index reads 12 and you feel calm, or reads 88 and you feel anxious, the divergence is informative about your positioning. Most people discover they feel exactly what the number says, which is itself the finding: you are not observing the crowd from outside it.

And it is a decent label for reviewing past decisions. Pull up your own buys and sells and note what the index read on each date. That is a cheap, concrete way to find out whether you consistently add into greed and trim into fear, which is a pattern you can address once you have seen it in your own record rather than in the abstract. The portfolio analytics that matter are mostly of this kind, measuring your behaviour rather than the market's.

What it is not good for is a decision. If a reading would change what you hold, the thing to interrogate is why a number substantially derived from last month's price volatility has standing to move your position when the price chart itself did not. Sentiment is context, and context belongs alongside a macro frame rather than alone: whether Bitcoin is moving with equities or against them tells you more about the character of a drawdown than any single-asset gauge, which is what Bitcoin against Gold and the S&P 500 is for.

The most useful discipline is to stop checking it daily. A number that updates every 24 hours and only means something at extremes is a number you can look at monthly without missing anything, and checking it more often mostly manufactures the feeling it claims to measure. That is the same problem as alert fatigue, and it responds to the same treatment.