Koryu · 黄龍
Koryu
← Research library
Research · 研究 · 42 · Foundations

Three majors up, eighty alts bleeding: breadth is how you know.

29 Jul 20264 min readFoundationsKoryu Research

This is an explainer of a participation metric, including the one this site computes daily. It predicts nothing. Nothing here is investment advice. Decisions are yours.

Market breadth measures how many assets participate in a move rather than how far the headline index travels: the share of a universe above its own trend line, advancing on the day, or making new highs. A market where three majors rise while eighty alts bleed and one where the whole board climbs can print identical headline returns, and breadth is the number that tells them apart.

The standard measurements

Breadth is a family of ratios over a defined universe, and how that universe is defined matters as much as the formula, so our universe rules are published in full. The workhorse: the percentage of assets trading above their own moving average, with the 50-day the common horizon, reading participation in the medium trend. Cousins: advance-decline counts (how many rose today), new-high and new-low counts over a window, and the share outperforming the reference asset, the version that turns alt season from a mood into a measurement. All are computable from public prices with published formulas, which is why breadth is one of the few crypto metrics that cannot be manufactured by a venue or a marketing department: it is arithmetic over the whole board, and the whole board is public.

Why breadth matters doubly here

In equities breadth is a useful nuance. In crypto it is closer to essential, for one structural reason: the headline of this market is a handful of majors, and the one-factor correlation regime means BTC's own strength can carry the visible market while the rest of the board quietly dies underneath. Headline-watchers cannot see that state; breadth is that state, measured. It is also the honest divergence detector in both directions: majors grinding up on collapsing breadth describes the 2024-2025 era of BTC dominance, precisely the tape where alt exposure bled against a rising market, while broad participation confirms that strength is a market phenomenon rather than an index artifact.

The classic readings

Three configurations carry most of breadth's information. Confirmation: rising prices with rising breadth, the boring healthy state. Negative divergence: rising headline prices over deteriorating breadth, historically the anteroom of trouble, because a market held up by fewer and fewer names has fewer and fewer reasons to stay up; this is the state the breadth floor in our regime dial's methodology exists to catch, and RISK-OFF triggers on breadth below 25% regardless of what BTC itself is doing. And the breadth thrust: from a washed-out floor, the share of assets reclaiming their trend lines jumping broadly and fast, historically among the more reliable early marks of regime turns, in crypto as in equities, because turns built on wide participation have the recruitment a durable trend requires. Thrusts are rare, which is their value; most bounces recruit narrowly and fail.

How our board computes it

The dial's breadth input is deliberately the plainest version: the percentage of the monitor's top-100 universe closing above its own 50-day average, computed daily from closed bars, thresholds fixed at 40% and 25% for the dial's state logic, universe membership published, every input recomputable against the monitor's published methodology. No smoothing, no proprietary weighting, no discretion. The reading appears alongside its raw value rather than as a verdict alone, so a reader can watch the number approach a threshold instead of being surprised by a state flip.

Honest limits

Four, stated plainly. Universe dependency: breadth over the top 100 liquid assets describes the market that matters for most readers, and says nothing about the ten-thousand-token long tail, a scoping choice, published as such. Equal-weight arithmetic: every asset counts once, so breadth deliberately ignores size, which is its point and its bias. Noise at the margin: thin alts oscillate around their averages, which is why the dial uses bands rather than a single line, and why day-to-day breadth wiggle means little against week-scale direction. And lag at turns: breadth built on 50-day averages inherits the same lag as every trend measure in this library; it will confirm regimes, not predict them. Breadth answers one question with unusual honesty, who is actually participating, and in a market this narrative-driven, that one honest answer earns its place on the board. Decisions are yours.

Related reading
FoundationsPosition sizing in crypto: surviving 8% daily noise5 min readFoundationsCrypto market sectors: L1s, L2s, DeFi, memes, rotation4 min readFoundationsThe halving cycle: evidence, narrative, and what's left of it5 min read
Frequently asked

What is market breadth?

A participation measure: the share of a defined universe above its own trend line, advancing on the day, or making new highs, rather than how far the headline moved. It distinguishes a market carried by three majors from one where the whole board climbs.

Why does breadth matter more in crypto?

Because the visible market IS a handful of majors, and the one-factor regime means BTC's strength can carry headlines while the rest of the board dies underneath. Breadth is that hidden state, measured, and it cannot be manufactured by any venue: it is arithmetic over public prices.

What is a breadth divergence?

Rising headline prices over deteriorating participation, historically the anteroom of trouble: a market held up by fewer names has fewer reasons to stay up. The 2024-2025 dominance era, majors grinding up while alt breadth collapsed, is the canonical recent case.

What is a breadth thrust?

From a washed-out floor, the share of assets reclaiming trend lines jumping broadly and fast, historically among the more reliable early marks of durable regime turns because wide participation is the recruitment a trend requires. Rare, which is the point; narrow bounces fail.

How does this site compute breadth?

The plainest version: the percentage of the top-100 universe closing above its own 50-day average, daily from closed bars, thresholds fixed at 40% and 25% in the regime dial's logic, universe published, every input recomputable, raw value shown alongside the state.