Ten lenses, weighted equally
Each lens asks one question about the market and answers it from our own published charts and price data. Every input is scored against its own readings of the past two years, so a lens reads 50 when its inputs sit mid-range and nearer 100 the hotter they run. The first seven read liquidity, flows, holders, leverage, rotation and the cycle; the last three read the price itself. The score is their average, each counting for 10%.
- 10%
Macro liquidity
Is the dollar liquidity that feeds risk assets growing or shrinking?
- 10%
Capital flows
Is new money arriving through the spot ETFs and stablecoins, and are coins leaving exchanges?
US spot ETF net flows, 30-day sum; Stablecoin supply on Ethereum, 30-day change; Exchange netflow, 30-day sum (withdrawals score higher)
- 10%
Holder behaviour
Are recent buyers in profit, are coins being sold at a profit, and are long-term holders selling?
Short-term holder MVRV; SOPR, 7-day average; Long-term holder supply change (selling scores higher)
- 10%
On-chain capital
Is capital flowing into bitcoin at the prices coins last moved, and how much of the supply is in profit?
Realized capital, 30-day change; Share of supply in profit
- 10%
- 10%
- 10%
- 10%
Price trend
Is the price above its long average, and is the 50-day above the 200-day?
Price against its 200-day average; 50-day average against the 200-day average
- 10%
Momentum
How strong has the last month's move been?
30-day price change
- 10%
Market structure
How close is the price to its yearly high and to the top of its 90-day range?
Distance below the 365-day high; Position in the 90-day range
Today's ten lens scores are in the daily read above only as the ones that lead and lag; their values and history are the Plus chart Market Regime Score: Lenses.
Ask about this page
Our AI analyst answers questions about this page from CryptoKeyhole’s own data and charts. Free with an account.
Asking opens a free account in a few seconds (a username and a password, no email); then the answer appears here.
The published record
| Day | Score | Band | Published |
|---|---|---|---|
| 7 November 2026 | 44 | Neutral | 00:08 UTC on 8 November 2026 |
At each day's close (00:00 UTC) the score goes into an append-only record exactly as it was published, with the method and the weights it was computed with. A row is never revised, even when an input is corrected later, and a day we failed to publish stays empty rather than being filled in afterwards. The record starts on 7 November 2026, the day the score was first published.
Everything before that date, on this page and in the chart, is back-calculated: today's method and weights applied to the inputs as they stand now. It shows how the score reads the past; it was not published at the time, and it is never called the record. The free API serves the record's latest 30 days, each row marked "basis": "published"; the Plus chart draws the whole record as its own line, beside the back-calculated one.
What followed past readings (back-calculated)
Back-calculated, the strongest 90 days followed risk-on readings (median +23%, 68% of windows higher) and the weakest followed defensive readings (median −9%). Overheated readings were followed by a gain over 90 days in the median, less than after risk-on ones: a hot score has meant a strong market more often than an imminent top.
| Band | Days | 30 days later (median) | 90 days later (median) | Higher after 90 days |
|---|---|---|---|---|
| Risk-off | 186 | +3% | −6% | 35% |
| Defensive | 1,054 | −1% | −9% | 42% |
| Neutral | 1,223 | +6% | +16% | 70% |
| Risk-on | 1,321 | +5% | +23% | 68% |
| Overheated | 164 | +2% | +20% | 65% |
Consecutive days share most of their next 90 days, so the day counts overstate how many independent cases there are, and one bull or bear market can fill a whole row. The published record is too short for this table yet; once it is long enough it will get its own, kept apart from this one.
How to read the score
- 80–100 Overheated Nearly everything runs hot at once: the price far above its averages, holders sitting on large profits, leverage stretched and money pouring in. Last back-calculated in this band: 17 December 2024.
- 60–79 Risk-on Most lenses lean hot: a rising market with money coming in and holders in profit. Last back-calculated in this band: 22 September 2026.
- 40–59 Neutral The lenses sit mid-range or pull against each other: no broad lean either way. Today's band, since 23 September 2026 (back-calculated before 7 November 2026).
- 20–39 Defensive Most lenses lean cold: a falling or stalled price, money leaving and holders under pressure. Last back-calculated in this band: 18 August 2026.
- 0–19 Risk-off Nearly everything is cold at once, as in the deepest stretches of bear markets. Last back-calculated in this band: 30 June 2026.
The score describes conditions, not timing. It is relative: every input is ranked against its own past two years, so after a long bull market it takes more to read hot, and after a long bear market less. A reading can stay in one band for months, and the bands are labels for ranges of the score, not signals.
The daily read names the two lenses running hottest and the two running coldest when they stand apart from the middle, and says when a reading is the highest or lowest for at least three months.
Use the score
The score is free to reuse under CC BY 4.0, credited to CryptoKeyhole with a link to this page. Today's reading and the published record are served without a key at /api/v1/market-regime (JSON, or ?format=csv for the record), the same answer for everyone, refreshed every ten minutes. The lens scores and the back-calculated history are not in it; they are Plus charts.
Lens scores and history with Plus
Two Plus charts carry the rest: Market Regime Score, with the published record and the back-calculated line drawn apart from October 2015, and Market Regime Score: Lenses, the ten lens scores beside the score. Signed out, both open on their 2019 to 2022 sample.
Sources and method
Inputs, all from CryptoKeyhole's own published charts and the site's admitted BTC/USD daily closes: Fed net liquidity's change over 90 days; the US spot bitcoin ETFs' net flows summed over 30 days, the 30-day change in stablecoin supply on Ethereum and the exchange netflow summed over 30 days, withdrawals scoring higher; short-term holder MVRV, SOPR averaged over seven days and the long-term holder supply change rate, selling scoring higher; the 30-day change in realized capital and the share of supply in profit; the leverage gauge; the altcoin season index; the cycle composite; the price against its 200-day average and the 50-day against the 200-day average; the 30-day price change; the distance below the 365-day high and the position in the 90-day range.
Each input is ranked against its own readings of the trailing 730 days, itself included, once there are at least 365 of them: 100 × (readings below + (equal readings − 1) ÷ 2) ÷ (readings − 1), turned round where lower means more risk-on. The leverage gauge, the altcoin season index and the cycle composite are already scored 0 to 100 against their own history, and the position in the 90-day range is 0 to 100 by construction, so these four are used as they are. A reading stands for the days after it for up to three days (ten for Fed net liquidity, five for the ETF flows, which report on business days). A lens is the average of its inputs with a reading; the score is the average of the lenses with a reading, each weighted 10% and renormalised over them, and a day needs seven of the ten. The back-calculated history therefore starts on 19 October 2015 with seven lenses; cycle position joins on 14 July 2016, altcoin rotation on 28 February 2018 and leverage on 2 September 2020.
The score is recomputed every ten minutes. The open day uses the latest hourly BTC price and each input's latest reading, so its score moves through the day; the closing score is fixed once the day's closing price is in and is then added to the published record, with its method version (ckh-market-regime-v1) and weights (equal-tenths-v1). The BTC valuation charts briefly hold while a new hourly quote is repriced; a run that meets the hold uses their last good reading and the record row lists any input read that way.
Historical daily prices: The Bitcoin Time Machine, CC BY 4.0; quotes since 3 October 2026: Alternative.me.
Questions
What is the Bitcoin market regime score?
A daily reading from 0 to 100 of how risk-on the Bitcoin market is, made by CryptoKeyhole from ten lenses weighted equally: macro liquidity, capital flows, holder behaviour, on-chain capital, leverage, altcoin rotation, cycle position, and the price's own trend, momentum and structure. Each input is scored against its own past two years, so 50 means conditions in the middle of their recent range.
Is a high score a signal to sell?
Why is the published record kept apart from the history?
Because a history recomputed today knows things that were not known at the time: inputs get corrected, late figures arrive and a method can change. The published record keeps each day's closing score exactly as we published it, from 7 November 2026, and is never revised; anything earlier is back-calculated with today's method and weights and is labelled that way on this page, in the chart and in the API.
How often does the score update?
Every ten minutes. The open day's score follows the hourly BTC price and the inputs as they arrive; the closing score is fixed at 00:00 UTC once the day's closing price is in, and goes into the published record.
Where does the data come from?
From CryptoKeyhole's own published charts: our Bitcoin node for holder behaviour, realized capital, supply in profit and exchange flows; our Ethereum node for stablecoin supply; the Federal Reserve's published balance sheet figures for net liquidity; the US spot ETFs' published holdings for their flows; futures market data for the leverage gauge; and our price data for the trend, momentum and structure lenses.
Open the Market Regime Score chartAll free tools
The score is calculated by CryptoKeyhole from its own published charts and price data and refreshed automatically. Nothing here is investment advice.