Cost Basis Phases

What is it?

Every Bitcoin holder bought their coins at some price - their cost basis. On-chain data lets us compute the average cost basis of three populations: Short-Term Holders (STH, coins moved recently), Long-Term Holders (LTH, coins dormant for a long time), and the network as a whole. Cost Basis Phases turns these three levels plus the spot price into a deterministic 8-state phase machine. The regime comes from one comparison: when the STH cost basis sits ABOVE the LTH cost basis, recent entrants are paying more than old holders - the market is in an expansion regime. When it sinks below, the market is in a compression regime. Within each regime, the position of the spot price inside the cost basis fan (above all three levels, between them, or below) defines four sub-phases - eight in total. A simple ordinal score adds the three comparisons: (price ≥ STH basis ±1) + (price ≥ network basis ±1) + (price ≥ LTH basis ±1), ranging from -3 (below everything) to +3 (above everything). There are no weights, no smoothing constants, no hidden parameters: the entire classification is reproducible from four public numbers per day.

How to read

The main panel shows the BTC price line on a logarithmic scale, with a colored area painted underneath it - the color IS the phase of that day. Three dashed lines frame the price: the STH cost basis (yellow), the network cost basis (orange) and the LTH cost basis (fuchsia). The lower panel renders the ordinal score (-3 to +3) as a histogram colored by the same phase palette. The header displays the CONFIRMED phase (a transition must persist 5 consecutive days before it is confirmed - this filters out flickering), how many days the market has spent in it, and the live distance in percent between the price and each of the three boundaries. Hovering any day updates the whole readout. Small markers above and below the price flag historically extreme days: green circles for realized-profit spikes, red circles for realized-loss spikes, green/red squares for profit/loss dominance, yellow circles for unusually low sell-side risk, and arrows for the rare STH×LTH crossovers that flip the regime.

Key zones

Expansion regime (STH basis above LTH basis): • Expansion (green, score +3): price above all three cost basis levels - every cohort is in aggregate profit, the trend self-reinforces. • Expansion · Strained (yellow): price slips below the STH basis - recent entrants are underwater while the broader market still holds; historically where mid-expansion corrections live. • Expansion · Waning (orange): price below the network basis - the average network participant is underwater despite an expansion regime; pressure builds. • Expansion · Broken (red, score -3): price below all three levels while STH basis still tops LTH - violent corrections inside a structurally expansionary ordering (March 2020 is the canonical example). Compression regime (STH basis below LTH basis): • Compression (violet, score -3): price below everything - capitulation territory, where the deepest value zones of past cycles formed. • Compression · Recovering (sky): price reclaims the STH basis - recent buyers are back in profit, the first structural improvement. • Compression · Maturing (cyan): price reclaims the network basis - the average participant breaks even; late-compression transition zone. • Compression · Breaking (fuchsia, score +3): price above all three levels while the regime is still compressed - historically the launchpad of new expansion regimes, resolved when the STH basis finally crosses back above the LTH basis.

What to observe

• Regime flips: the STH×LTH crossover happens roughly twice per cycle and is the structural pivot of the whole framework. Watch the arrow markers - a bull flip after a long compression has historically marked the early innings of new expansions. • Confirmed vs raw phase divergence: when the header shows a confirmed phase but the area color disagrees for a few days, a transition attempt is underway. Repeated failed attempts around one boundary reveal a contested level. • Days-in-phase: phases that overstay their historical norms (visible by comparing colored stretches across cycles) often precede regime transitions. • Distance compression: when the three dashed lines converge and the price weaves between them, transitions multiply - the fan tightening is itself an information event. • Event clusters: realized-loss spikes plus loss dominance inside a Compression phase have historically clustered near capitulation lows; realized-profit spikes plus profit dominance inside an Expansion phase near euphoric peaks. A low sell-side risk reading layered on a Compression or Compression · Recovering phase characterizes exhausted sell pressure.

Historical context

Across the full history since 2011, the phase machine recounts every cycle in the same grammar. The 2013 double-top produced a rare cluster of rapid regime flips (September-October 2013) - a genuinely chaotic period where the STH and LTH cost basis braided around each other. The 2015 and 2018-2019 downturns spent their longest stretches in the violet Compression phase, with the loss-dominance and low sell-side-risk markers concentrating near the December 2018 capitulation. March 2020 (COVID crash) registered as Expansion · Broken rather than Compression: the collapse was too fast for the cost basis structure to invert, which is precisely why the recovery was so explosive. The November 2022 trough (post-FTX) painted the textbook sequence: Compression with loss-dominance clusters, then Compression · Recovering in early 2023, Compression · Maturing through the autumn rally, and an expansion regime flip in December 2023 before the ETF-driven expansion. Each cycle since 2011 has spent roughly a third of its days in the green Expansion phase and a quarter in the violet Compression phase, with the six transition phases sharing the remainder.

Expert notes

Methodological conventions, locked for reproducibility. (1) Hysteresis: the confirmed phase only transitions after the raw classification persists 5 consecutive days; on confirmation the days-in-phase counter is credited retroactively with the streak. This cuts transition noise by roughly two-thirds without altering the raw record, which remains fully painted in the area. (2) Extreme-day markers use order statistics computed strictly on past data - the threshold for any day never includes that day's value. Realized profit and loss spikes use an expanding window at the 99.75th percentile; profit-to-loss dominance uses a rolling 4-year window at the 99.75th/0.25th percentiles; low sell-side risk uses a rolling 4-year window at the 0.5th percentile. No marker fires during the first 4 years of available history, which is why the early period shows none. (3) The framework is a public-domain convention built on industry-standard realized price cohorts - it is intentionally NOT proprietary, and its value lies in its transparency. (4) The STH/LTH cohort boundary follows the established industry convention based on coin age. (5) The ordinal score is a position summary, not an intensity measure: +3 says the price is above all three levels, not how far above. (6) The cost basis variant used here is the time-weighted (capitalized) form of the cohort cost basis - each coin's acquisition price is weighted by holding time. This is what makes the STH×LTH crossovers structurally meaningful, and it intentionally differs from the classic realized price cohorts (realized cap divided by cohort supply, without time weighting), which live in their own dedicated chart. (7) Acknowledgement: the cost-basis phase reading was popularized by the work of Nym21; this implementation is an independent Trinity build of that public-domain idea, with its own Expansion/Compression vocabulary, hysteresis, ordinal score and extreme-day event layer.

Common mistakes to avoid

• A phase is a position, not a forecast: 'Expansion · Waning' describes where price sits in the cost basis fan today - it does not promise the expansion is ending. Phases have reversed at every boundary in history. • Expansion · Broken is not a compression regime: the regime ordering still favors expansion; conflating the red phase with the violet Compression phase misreads the structure (March 2020 resolved upward within weeks). • Regime flips can cluster: during chaotic tops the STH and LTH basis can braid, producing several flips in weeks (autumn 2013). A single flip is not always a clean, durable pivot. • The score is ordinal, not linear: moving from +1 to +3 is a change of position, not 'twice as expansionary'. Distances to the boundaries (shown in the header) carry the magnitude information. • Extreme-day markers are descriptive statistics, not recommendations: they flag percentile rarity of on-chain flows, nothing more. • The colored area uses the RAW phase by design: short color flickers around boundaries are real data, filtered only in the confirmed header readout - not hidden.

Programmatic access

REST API

curl -sS \
  'https://api.trinityinsights.io/api/v1/cycle-intelligence/cycle-cost-basis-phases/history?days=90' \
  -H 'X-API-Key: $TRINITY_API_KEY'

MCP server

{
  "tool": "get_chart_value",
  "metric_id": "cycle-cost-basis-phases",
  "timeframe": "1y"
}

Required tier: pro. See the pricing grid for the tier list and the MCP documentation for multi-client configuration.

Related metrics

• cycle-cost-basis-thermometer
• cycle-phase-probability
• cycle-quantile-zones-cold-warm-hot
• cycle-pnl-vs-ma365-regime
• cycle-emotion-roadmap

Institutional disclaimer

Trinity Insights is an educational and analytical tool. The metric above does not constitute investment advice. Trinity Insights is not a Crypto-Asset Service Provider (CASP) registered under MiCA Regulation (EU) 2023/1114. See the full disclaimer.