Bitcoin Rainbow Chart (Live)
At $71,460 (close of Nov 12, 2026), Bitcoin sat in the “Very Cheap” band of the rainbow chart — the 22th percentile of its entire price history relative to trend. Unlike the classic frozen 2014 model, this chart refits live (at most once an hour) on all 17 years of data, and its bands are calibrated to the measured spread of that history — which is why they narrow over time.
Band names describe past price levels, not advice. Data through Nov 12, 2026 (daily close, UTC).
By Menno van Ravels - 13 years in crypto, 3 bear markets survived, zero paid promotions
The Excursion Ledger
Every time in 17 years that Bitcoin closed outside the entire rainbow — episodes under the current fitted model (refitting shifts old boundaries slightly). What happened afterwards is history, not a forecast.
- Duration:
- 19 days
- Max distance:
- -2.8%
- 90 days after it ended:
- BTC +63.3% (by Apr 8, 2023)
- Duration:
- 10 days
- Max distance:
- +11.9%
- 90 days after it ended:
- BTC -47.7% (by Jul 16, 2021)
- Duration:
- 9 days
- Max distance:
- +4.7%
- 90 days after it ended:
- BTC -41.9% (by Jul 5, 2021)
- Duration:
- 12 days
- Max distance:
- +10.1%
- 90 days after it ended:
- BTC -38.1% (by Jun 19, 2021)
- Duration:
- 7 days
- Max distance:
- -1.7%
- 90 days after it ended:
- BTC +77.0% (by Jan 3, 2016)
- Duration:
- 9 days
- Max distance:
- -3.3%
- 90 days after it ended:
- BTC +78.5% (by Dec 26, 2015)
- Duration:
- 7 days
- Max distance:
- +89.3%
- 90 days after it ended:
- BTC -72.7% (by Sep 8, 2011)
Plus 28 shorter blips of under a week (totalling 53 days), mostly in the volatile early years. Not a forecast. Past excursions ended in different market regimes; the sample is tiny by construction.
The Honest Version of This Chart
The Bitcoin Rainbow Chart started in 2014 as a joke on Reddit, mocking people who drew arbitrary trend lines on Bitcoin’s price. It has no scientific basis, and the classic version has two structural problems: it fits a straight line through log-log space even though Bitcoin’s growth flattens every cycle, and its colored bands sit at offsets someone hand-picked in 2014 and never re-measured. That is why, on every frozen rainbow site, the bands drift further above the actual price each year.
We do two things differently. First, the trend is refit live on the full daily history with a decaying-growth term, so the curve is allowed to flatten the way the data actually does. Second, the 9 bands are not hand-picked: each band edge is a quantile curve fitted to the same history, so “Fire Sale” literally means “the cheapest few percent of Bitcoin’s recorded life relative to trend” — and because Bitcoin’s volatility has compressed cycle after cycle, the bands narrow over time. The rainbow shrinks. Use the “+ Classic 2014” toggle above to see how far the frozen recipe has drifted from the data.
None of this makes the chart predictive. It is a rear-view mirror: a description of where today’s price sits inside 17 years of history, which shifts a little every time it refits (we publish that drift below). We refuse to quote an accuracy percentage for it, and you should distrust any rainbow chart that does. Not investment advice.
How the Bands Are Calculated (Check Our Math)
Model: quantile regression of ln(price) on [1, x, x²] where x is normalized ln(days since the genesis block), fitted on 6,156 daily closes through Nov 12, 2026. Ten boundary curves at the 1 / 5 / 15 / 30 / 40 / 60 / 70 / 85 / 95 / 99th percentiles form the nine bands; below the 1st / above the 99th percentile curve is open space — the chart never paints a band there, and never draws bands into the future. Monotone rearrangement resolves 3 raw curve crossings in the sparse pre-2011 region.
| Zone | Claimed share | Measured (all-time) | Measured (last 4y) | Out-of-sample (2019+) |
|---|---|---|---|---|
| Below the rainbow | 1% | 1.0% | 1.8% | 3.5% |
| Fire Sale | 4% | 4.0% | 5.2% | 4.5% |
| Very Cheap | 10% | 10.0% | 10.6% | 5.9% |
| Cheap | 15% | 15.0% | 18.5% | 9.7% |
| Still Cheap | 10% | 10.0% | 10.0% | 8.5% |
| Around Trend | 20% | 20.0% | 19.4% | 25.7% |
| Is This a Bubble? | 10% | 10.0% | 11.6% | 13.0% |
| FOMO Intensifies | 15% | 15.0% | 15.6% | 15.9% |
| Overheated | 10% | 10.0% | 6.7% | 7.6% |
| Maximum Bubble Territory | 4% | 4.0% | 0.6% | 3.3% |
| Above the rainbow | 1% | 1.0% | 0.0% | 2.4% |
The honest numbers, in full: on a typical day, price closes within ×1.56 of the trend curve (up or down). Walk-forward — refit monthly on only past data, then scored on the 2,768 out-of-sample days since 2019 — the median absolute gap between price and the trend was $11,601. The gap between claimed and measured shares in the table is the model’s real calibration error; we publish it instead of hiding it. Refit disclosure: the bottom boundary moved 0.0% over the past 30 days of refits. We deliberately do not report R² (always looks near-perfect on a trending log series), accuracy percentages, confidence intervals (invalid under autocorrelation), price targets, or backtested returns.
What the Bands Mean
Reading bottom to top, each band is a measured claim about 17 years of history, not a vibe: Fire Sale is the 1st-5th percentile of price-vs-trend (BTC has closed in or below it roughly 5.0% of all days), Very Cheap and Cheap cover the cheap quarter, Around Trend straddles the trend itself, and Overheated with Maximum Bubble Territory mark the most expensive few percent — where the December 2017 and November 2021 peaks lived. When price leaves the rainbow entirely, it enters territory seen on well under 1% of days; the Excursion Ledger above shows the sustained (week-plus) episodes and summarizes the shorter blips.
For a cycle framework that doesn’t rely on curve-fitting at all, see the Bitcoin Cycle Chart, pair this page with the free Bitcoin risk score, or backtest a plan in the DCA simulator.