This article describes historical phases and their signatures. Phase boundaries are obvious only in hindsight, and this piece claims no ability to call them in real time. Nothing here is investment advice. Decisions are yours.
The crypto market cycle is the observed rhythm of this asset class: quiet accumulation after a collapse, an advance that broadens into mania, a top that distributes, and a winter that erases most of the move, repeating with variations every few years. The phases are real in the record. Their edges are visible only backwards.
The four phases, on the actual tape
Accumulation: the post-capitulation flat, volatility compressed, volume thin, attention gone; 2015 after the first great winter, 2019 after 2018, early 2023 after 2022. Markup: trend arrives, first in BTC, then breadth broadens down the risk curve as confidence recruits; 2016-17, 2020-21. Mania is markup's terminal stage: vertical price, universal participation, junk leading quality, leverage everywhere; late 2017, and late 2021's twin peaks. Distribution and winter close the loop: tops that churn while breadth quietly dies, then declines that run to the drawdown base rates this library documents, roughly -85% from the 2013 peak, -84% from 2017, -77% from 2021 for BTC itself and far deeper down the tail, each winter lasting years rather than quarters.
The wheel's mechanics
Why the shape recurs. Reflexivity: price is this market's marketing, so rising prices recruit buyers whose buying raises prices, a loop that runs identically in reverse, making both directions overshoot. Leverage: each advance rebuilds the borrowed layer, and each turn flushes it through cascades that convert corrections into collapses. Narrative rotation: capital moves through story families in sequence as the cycle matures, majors to sectors to pure junk, so where the froth lives is itself a cycle clock. And the halving calendar supplies the coordination myth, real mechanism, shrinking magnitude: whatever its causal residue, enough participants schedule around it that the schedule participates in the cycle it claims to predict.
Phase signatures a board can measure
Cycle talk earns its keep only when it cashes out into measurable state, which is what this site's instruments exist for. Accumulation prints as compressed ATR percentages, dead momentum columns, and breadth flat at low levels. Markup prints as trend states stacking positive on the moving averages, and breadth climbing through its thresholds. Mania prints as extreme momentum readings concentrated in the board's thinnest names while relative strength versus BTC broadens indiscriminately. Distribution prints as the divergence this library keeps returning to: headline prices holding while breadth decays. None of these signatures is a timing device; together they are a where-are-we instrument panel, which is the honest ceiling of cycle analysis and exactly what the dial methodology is built to report.
The honest limits
Three, stated against the folklore. Boundaries are hindsight: every phase label is assigned confidently years later and contested bitterly in real time, which is why this site publishes regime state, computable today, rather than cycle position, arguable forever. The sample is tiny: a handful of full cycles is anecdote-grade evidence for any precise periodicity claim, the same n-of-four problem the halving carries. And the script mutates: the 2024-2026 era broke choreography that held for a decade, new highs before the halving, institutional flows through ETFs replacing the retail mania engine, majors strength with alt breadth notably absent, and any cycle model that cannot survive its script breaking was a costume, not a model. The cycle is a useful map of how this market has breathed. The dial exists because maps age and measurements don't. Decisions are yours.