Sector labels describe narrative families, not audited industry classifications, and this article treats them with corresponding skepticism. Nothing here is investment advice. Decisions are yours.
Equities have sectors because companies have industries. Crypto has sectors because tokens have STORIES, and the difference explains both why crypto sector analysis works, stories move capital in herds, and why it disappoints anyone expecting equity-grade structure underneath. Here is the working taxonomy, what it actually predicts, and where it breaks.
The working taxonomy
The labels that organize most sector data. Layer 1s: base blockchains, the market's blue-chip complex after BTC itself. Layer 2s: scaling networks settling to a base chain, a family whose fortunes track their parent ecosystems. DeFi: exchange, lending, and derivatives protocols, the family with the most measurable fundamentals, fees and usage exist here, and a history of its own boom in 2020-2021. Memes: assets whose explicit value proposition is attention, the purest expression of crypto as a coordination game, and in recent cycles among its highest-volume families. Infrastructure: oracles, data, storage, interoperability. Gaming and NFT ecosystems. AI tokens: the newest large family, capital expressing an equity-market theme through crypto wrappers. And real-world assets, tokenized traditional instruments. The boundaries blur, projects claim multiple memberships, and aggregators disagree; the taxonomy is a map of narratives, which is exactly why it works as well as it does.
What sector membership actually predicts
In equities, sector membership predicts shared cash-flow exposures. In crypto, it predicts shared ATTENTION exposure: when a narrative catches, capital floods its whole family more or less indiscriminately, the strongest names first, the barely-related hangers-on shortly after. This makes sector co-movement in crypto FASTER and less discriminating than its equity cousin: a DeFi headline lifts fifty tokens by dinnertime, most of which share nothing but the label. For a momentum reader this is genuinely useful information: strength that appears across a family simultaneously is a narrative arriving, historically more durable than a single name's spike, which is why leaders clustering within a hot family has been a classic continuation tell, while a lone mover in a dead family invites the pump suspicion, per the Telegram assessment.
The diversification illusion, again
The equity intuition, spreading across sectors spreads risk, mostly fails here, and the failure has a number. In calm tapes, crypto sectors do differentiate: memes and majors can run opposite directions for weeks, and family dispersion is real. In stress, correlations converge toward one across EVERY family, because the marginal seller is not choosing between narratives, they are exiting the asset class, and the one-factor structure documented across this library, detailed in the dial methodology, flattens the taxonomy precisely when protection was the point. Sector spreading in crypto diversifies your BULL market experience and your narrative exposure; it does almost nothing for your winter, a distinction our founding experiment paid to relearn with position-count diversification, as documented in the full transplant experiment.
Reading sectors for momentum purposes
Three practical readings survive the caveats. Family breadth: when a sector's members broadly clear their trend lines together, that is narrative arrival with participation, the sector-level cousin of the market breadth reading in the breadth explainer. Leadership structure: healthy family moves are led by the family's most liquid names, junk-first rallies within a sector have historically been late-stage or manufactured. And rotation tracking: capital migrating family to family, visible as one sector's breadth fading while another's ignites, is the medium-term texture of crypto bulls, the finer-grained version of the majors-to-alts rotation story, per the rotation piece, and carries the same warning label: a tendency, never a timetable.
Data-quality caveats
Sector analytics inherit every data problem this library documents, plus one of their own: membership is editorial. Aggregators hand-assign categories, disagree with each other, lag new narratives by months, and let projects self-describe into hot families, AI in particular collected tokens whose connection to the theme is a press release. Any sector index is therefore a curated basket with the curator's judgment embedded, unpacked in the index piece, its volume figures carry the wash-trading pollution of thin constituents, laid out in the wash-trading piece, and its market-cap weighting imports the FDV fictions of low-float members, per the FDV explainer. Our own board displays per-asset categories as descriptive metadata and builds nothing predictive on them, treating sector labels the way this article recommends: as a map of stories, useful for reading the crowd, unfit for bearing structural weight. Decisions are yours.