Charted valuation and flow indicators, read server-side from public sources. Every figure names its source and its parameters — these are model outputs, not facts. No buy or sell guidance, no position sizing.
7 indicators, each placed on the conventional threshold bands for that indicator and weighted 20/15/15/15/15/10/10. A higher score means the bands read cheaper. Ranked instead against their own daily histories, the same readings run from Fear & Greed 27 to Price ÷ 200-week line 79. The bands are a convention, not a measurement — each chip below carries its own reading.
score
Today
Showing the 2026-09-01 daily snapshot — live values replace these once the page loads.
Cycle radar and temperature
Two views of the same thing: where each indicator sits in its own history, not against a threshold someone published. The chips in Today's read carry these same percentiles; this is the shape of them.
These two are not the same measurement, and they can disagree. The score places each reading on the conventional bands for that indicator — thresholds the industry publishes, a convention. The temperature ranks each reading against its own daily history — arithmetic on our data. When the score reads cheap and the temperature does not, it means cheap by the old rules but ordinary by this indicator's own recent distribution. Neither is the truth; they answer different questions.
AHR999
AHR999 cycle valuation
2026 refit power law · a 5.5856 · b −16.1567 · where price sits against long-run cost and the power-law trend
The bands are the indicator author's own vocabulary — he defined 0.45 and 1.2; the 5 boundary is a later, widely-copied extension. None of it is guidance from this page, and the levels above are arithmetic, not entry points: they answer “if price were here right now, what would AHR999 read”, holding today’s cost and fit fixed. Both inputs drift daily.
AHR999 history
Daily readings against the two anchors, on the same fit as the numbers above.
Where it breaks: it assumes price reverts to a power law, so in a sustained one-way market it parks in one band for months. It is also acutely sensitive to which power-law fit you use — see the next section, where the two fits differ by about a third.
Power-law value
A fit of price against Bitcoin's age, and one of the two anchors AHR999 uses. The parameters this page uses are printed below — it is a fit, not a constant of nature, and refitting on a different window moves the answer.
The corridor: support, fair, resistance
The fair line is the same fit as above — the page does not carry a second power-law valuation, because two different fair values would only leave you guessing which to trust. Support and resistance are that same line shifted down and up by the 1st and 99th percentile of its own daily residuals, so all three share a slope and differ only in intercept. Every number below is reproducible from the parameters printed with it.
It is a statistical envelope, not a floor. The support line is drawn where 1% of past days closed below it — so by construction, price being under it is not impossible, it is the 1-in-100 case that defined the line.
If you came looking for a “floor model”, this support line is the one this page carries, and the paragraph above is why it is not called a floor. Other sites publish floor models built on third-party datasets under their own names; this page does not reproduce those, because it could not show you how they were computed. What it can do is print its own parameters, which are directly above.
The residual window is the last two halving epochs (2016-07-09 onward, 3,707 days, R² 0.9614), not the whole history. Before 2016 the market’s depth and participants do not resemble today’s, and including that era widens the corridor until it stops discriminating — on the full history the same percentiles give $38,504 and $930,644 instead. Halvings are Bitcoin’s own unit of time, which is a firmer reason to cut there than picking a year. You can see the size of that choice above; it is a choice, not a measurement.
Market structure
Price and its long-term trend on one axis — structure first, noise second. Halving dates appear as dashed markers once the range is long enough to contain one; on 90 days or a year there is none to show. The last halving was 20 April 2024.
The 200-day average is the arithmetic mean of the same closes the DCA cost uses as a harmonic mean — a fixed-sum daily buyer pays the harmonic one. The 200-week average spans roughly one halving cycle, which is why it gets read as a long-run floor; it is a four-year lookback, so it always turns late.
Mining
Miner shutdown band
—The price at which mining stops covering its own electricity. The newest rigs give up last, six-year-old rigs give up first.
Next halving
—The block height is exact. The date is not — it assumes the 10-minute target block time.
Cost basis and the MVRV band
The realised price is what the average coin last moved at — the market’s aggregate cost basis. It needs no extra source: MVRV is market cap over realised cap, so realised price = price ÷ MVRV. Today that is $53,081, and price sits 46% above it, which is another way of saying MVRV is 1.4583.
This is the aggregate, not a distribution. A per-UTXO cost basis — what each cohort of coins paid — needs UTXO-level chain data this page does not carry, so it is not what you are looking at. What is below is the price level MVRV would imply at each point of its own historical range, holding today’s cost basis fixed.
| MVRV at | Value | Price that implies | |
|---|---|---|---|
| 5th percentile | 0.823 | $43,700 | price is above this |
| 25th | 1.274 | $67,643 | price is above this |
| median | 1.695 | $89,989 | price is below this |
| 75th | 2.181 | $115,751 | price is below this |
| 95th percentile | 3.283 | $174,289 | price is below this |
On-chain
These three are daily, not live. They are computed from the previous UTC day’s chain data, so they do not move with the price on this page. Puell compares miner revenue against its own yearly average; Reserve Risk weighs price against long-term holder conviction; SOPR is above 1 when the coins moving that day are moving at a profit.
Flows and sentiment
Two cautions. The annualised funding figure extrapolates one 8-hour period and is not a return you can collect. The Coinbase premium also contains USDT's own deviation from $1 — when USDT trades at a discount the premium rises for reasons unrelated to dollar demand.
Market-implied odds
Current traded prices for these strikes on a prediction market, read by convention as probabilities. What bettors quote now — not a forecast, and not this page's view. Thinly traded strikes quote erratically.
Your position
Arithmetic on numbers you type, in your browser. Nothing is sent anywhere and nothing is stored on our side — the two values below live in this browser's local storage and you can clear them with the button. It is not advice and it is not a forecast: each column below holds the price fixed at a stated level and divides, which is all it does. Fees, spreads and taxes are not included.
Type a holding or a budget to see the arithmetic.
The calculator is free and stays free. If you would rather be told when a buy-side condition holds instead of checking, there is an email alert — one email per day at most, and the page prints how many that has actually been, year by year, before you decide.
Data health
Source, whether it was fetched, and how old the underlying data is. Nothing stale is substituted for a value that could not be fetched.
| Indicator | State | Source | Data as of |
|---|
- Price and daily closes: Coinbase Exchange daily candles, close of each bar. Window is — days, so any high or low here is within that window, not all-time.
- Change is the difference between the last two daily closes, not a rolling 24 hours.
- Weekly closes for the 200-week average: DefiLlama. Hashrate and difficulty: mempool.space. Fear & Greed: Alternative.me. MVRV: Coin Metrics community API. Funding: Bybit, falling back to OKX. Stablecoin supply: DefiLlama. Puell Multiple, Reserve Risk and SOPR: BGeometrics.