What is market sentiment?
The aggregate mood toward an asset, estimated by having a model read what people wrote about it.
The definition
Market sentiment is the collective attitude of participants toward an asset — broadly, whether the mood is positive or negative. It is a real thing people experience and not a directly observable quantity, so measuring it means finding something observable that stands in for it.
In practice that proxy is public text: posts, comments and headlines mentioning the asset, classified by tone and aggregated into a number.
How a number gets produced
- Collect posts mentioning the asset from covered platforms.
- Classify each one as positive, negative or neutral.
- Aggregate the classifications into a single score per asset per window.
On Moonboard the result is a sentiment field from 0 to 100, where higher is more positive. It feeds more metrics than any other single input — the Moonboard Score, OPS, COR, SDE and the retail leg of WSD.
Why crypto text is hard to classify
Irony is the default register
Large parts of crypto discourse are ironic by convention. Losses are celebrated, self-deprecation signals belonging, and enthusiasm is expressed through exaggeration that reads as sarcasm. "Great, another perfect entry" after a 30 % fall is sincere frustration, and reliably misread.
Vocabulary carries inverted valence
Terms that are negative in general English are positive in crypto usage, and vice versa. A classifier trained on product reviews or news has learned the wrong sign for a meaningful part of the vocabulary — and the vocabulary shifts fast enough that any fixed lexicon decays.
Communities rally under stress
The most consequential failure. During a crash, communities produce large volumes of resolute, forward-looking language, and classifiers read much of it as positive.
The bias therefore points in a specific direction: sentiment tends to overstate positivity precisely when prices are falling hardest. Any contrarian use of sentiment has to account for that, because the measurement error runs the same way as the effect being looked for.
Sentiment is not the same as a mood index
| Social sentiment | Composite mood index | |
|---|---|---|
| Built from | Text classification only | Text plus price action |
| Scope | Per asset | Whole market |
| Independent of price? | Yes | No, largely by construction |
Moonboard's market mood weights price more heavily than social, on the reasoning that communities post "buy the dip" at every level while price is harder to fake. That is defensible — and it means a price-weighted mood index is substantially a transformation of recent returns. It summarises what prices already show rather than confirming it independently.
Reading a sentiment score
- Check the volume behind it. A sentiment of 72 from 8 000 posts and one from 6 posts look identical downstream. Every metric treats them identically. Thin coverage means an indicative figure, not a measured one.
- Extremes are least reliable. Very high and very low readings are disproportionately produced by unusual text — crisis language, coordinated promotion, sarcasm — which classifiers handle worst.
- Extremes persist. Sentiment can sit near a bound for weeks. "Extreme" describes a state, not a turning point.
- Divergence is more informative than level. Sentiment disagreeing with an independently measured quantity is the interesting case — see sentiment divergence.
Coordinated and generated posts are classified and aggregated exactly like organic ones. Manufacturing positive sentiment for a small asset is inexpensive. See AI agents and synthetic activity.
Related
- Sentiment researchThe full treatment of classification and aggregation
- Sentiment divergenceComparing mood against market data
- SDE methodologyHow Moonboard implements a divergence measure
- Social momentumVolume of conversation, rather than its tone