Every one of these charts is presented as a way of knowing where Bitcoin is going. We rebuilt each of them from a single public price series, fitted them the way their authors do, and then did the one thing the charts are rarely subjected to: extended them forward without adjusting anything, and compared them with what happened.
Method
What we tested for
The same four questions were put to every chart. They are deliberately not about whether a chart has been right lately — a stopped clock manages that — but about whether it could ever be shown to be wrong.
Could it fail?
Is there an outcome the chart forbids? A construction compatible with every price carries no information about any of them. This is the question most of them fail first.
Does it survive out of sample?
We fitted each model on early data only, extended it without refitting, and compared it with what followed. Fitting the past is not a result; the past is what it was fitted to.
How many free parameters, how few events?
Window lengths, multipliers and cut-off dates are choices. When several of them are tuned against three or four historical peaks, a good fit is arithmetic, not discovery.
Does the mechanism do any work?
Several charts are narrated with a real economic story — scarcity, network growth, miner capitulation — while the arithmetic underneath uses only price and elapsed time. Where a mechanism exists, we tested whether removing it changes the result.
Findings
The verdicts
Two verdicts only, because a scale with more rungs would be a way of avoiding the question.
| Chart | What it is made of | Verdict |
|---|---|---|
| Bitcoin Rainbow Chart | A regression curve is fitted through the historical price on a log scale, and the bands are fixed offsets above and below that curve. Nothing enters the calculation except past price and elapsed time — no demand, no supply, no flows, no rates. The colours and their names are chosen by the author, not derived from anything. | No predictive value |
| Bitcoin Power Law & log regression bands | log(price) is regressed against log(time since the genesis block). The result is a straight line in log-log space, and bands are added as offsets or percentiles around it. The only inputs are price and elapsed time. | No predictive value |
| The four-year cycle | The halving dates are known in advance, so the model consists of counting months from each one and asserting that the same sequence repeats. The “lengthening cycles” variant adds that each cycle takes longer and returns less than the last. | Descriptive, oversold as prediction |
| Pi Cycle Top Indicator | Two simple moving averages, one multiplied by two, and the date on which they cross. The name comes from 350 divided by 111 being close to π — an observation about the chosen numbers, not a reason for choosing them. | No predictive value |
| Stock-to-Flow | The ratio is computed from Bitcoin’s known emission schedule, and market value is regressed against it in logarithms. The published relationship is ln(market value) = 3.3 × ln(S2F) + 14.6, reported with an R² of 95 per cent. Because the emission schedule is fixed in advance, that curve produces dated price targets years ahead — which is the whole appeal, and the reason the model can be checked. | No predictive value |
| Hash Ribbons | Two moving averages of network hash rate, and the dates on which one crosses the other. Unlike the price-only constructions elsewhere in this study, the input is a physical quantity: how much computation is securing the network. | Descriptive, oversold as prediction |
| Golden Ratio Multiplier | One moving average, multiplied by 1.6, 2, 3, 5, 8, 13 and 21. The multipliers come from the Fibonacci sequence; nothing in the market determines them. | No predictive value |
The charts
One page each, with the calculation
Bitcoin Rainbow Chart
The Rainbow Chart draws Bitcoin’s price on a logarithmic scale and lays coloured bands over it, from “fire sale” at the bottom to “maximum bubble” at the top. The colour under today’s price is presented as a verdict on whether Bitcoin is cheap or expensive.
Read the analysisNo predictive valueBitcoin Power Law & log regression bands
Plotted with both axes logarithmic, Bitcoin’s price since 2010 falls close to a straight line. The power-law reading treats that line as a law of growth and extends it forward, sometimes with bands around it, to say where price belongs in any given year.
Read the analysisDescriptive, oversold as predictionThe four-year cycle
Bitcoin is said to move in four-year cycles anchored to the halving: accumulation, a rally peaking roughly a year to eighteen months after the halving, then a deep bear market. The cycle is used to date tops and bottoms in advance.
Read the analysisNo predictive valuePi Cycle Top Indicator
When Bitcoin’s 111-day moving average crosses above twice its 350-day moving average, the indicator is said to mark the top of the cycle within days.
Read the analysisNo predictive valueStock-to-Flow
Stock-to-flow divides the existing supply of an asset by the amount produced each year. The model asserts that Bitcoin’s market value follows this scarcity ratio, so that each halving — which doubles the ratio — implies a step change in price.
Read the analysisDescriptive, oversold as predictionHash Ribbons
When the 30-day average of Bitcoin’s hash rate falls below the 60-day average and then recovers, miners are said to have capitulated; confirmed by a short-term price crossover, the recovery is read as a buy signal.
Read the analysisNo predictive valueGolden Ratio Multiplier
Bitcoin’s 350-day moving average is multiplied by a series of Fibonacci numbers, and the resulting bands are presented as levels at which cycle tops and intermediate highs occur.
Read the analysisDescribing is not predicting
Moonkelp publishes indicators too, and the distinction this study rests on applies to them as well. MVRV, the Mayer Multiple, dominance and the Altcoin Season Index describe a state that exists right now: what holders paid, how far price has run from its own trend, where capital sits. They can be high or low, and they are silent about next month.
None of them contains a curve fitted to the past and extended into the future, because that is the step where description turns into prophecy — and, as every page in this study shows, the step where the evidence stops.
Method & limits
All figures come from Coin Metrics Community Network Data — the Bitcoin reference rate and hash rate, daily, from 2010 through 14 August 2026. Nothing after that date is covered, and no claim on these pages depends on it. Moving averages, regressions and forward returns were computed directly from that series; each page states what was calculated and how.
Where the pages quote what an individual published — the 2019 stock-to-flow paper, its author’s December 2021 posts, the Rainbow Chart maintainer’s own disclaimer — every quotation carries a link to the primary source or an archive copy, and each is dated. The core of the study is still the harder test: a chart either survives an out-of-sample extension or it does not, regardless of who published it.
The stock-to-flow page held this study up the longest for exactly that reason: its central question is about a documented public record rather than about arithmetic, and it went live only once every quotation could be sourced first-hand.
This study is reviewed quarterly; corrections normally ship within days. Perspectives, not investment advice. Corrections and challenges are welcome: contact@moonkelp.com.