Power-Law Rainbow

Price against bands parallel to the power-law fit — a log-log rainbow corridor.

Data to 2026-09-22. On-chain metrics are fetched on a rotation, so this can trail the build by a few days.

How to read it

Bands drawn parallel to the long-run power-law fit, so each represents a constant multiple of the fitted trend. Price moving into the upper bands means it is stretched above its historical trajectory; the lower bands are where past cycle lows have occurred.

How it's calculated

The power-law trend is fitted by least squares to log price against log time since genesis. Bands are that fitted line shifted by fixed offsets in log space.

What it did in past cycles

Price has closed above the top 'Overheated' band on 151 of 4,345 days — about 3.5% of the record — and the last of those days was 2021-05-11. It has not happened since.

The three most recent cycle tops show the drift clearly:

2017-12-16 — price $19,497 against an Overheated band at $8,584: price was 2.3× above the top band
2021-11-08 — $67,567 against $70,433: just under it
2025-10-06 — $124,753 against $331,016: price reached only 38% of the top band

The bands have been rising faster than price does. A power law fitted through Bitcoin's steep early years extrapolates a trajectory that later cycles have not matched, so the whole rainbow drifts upward away from the market it is meant to describe. Anyone who bought the 'Fire sale' band in 2026 expecting the 2017 relationship would have been buying a line that has moved since.

What it doesn't tell you

The bands are descriptive, not predictive — they are drawn from the fit, so of course price has stayed within them. The band labels are conventional names, not valuations, and the entire structure shifts as new data extends the fit.

That last point is not a technicality. The fit is re-estimated on every build, so a band's dollar value changes over time and the label 'Overheated' has meant several different things. Screenshots of this chart from different years are not comparable, and neither are the conclusions drawn from them.

Reading it with other metrics

The power law page shows the underlying fit without the bands, along with what the fit actually is and how far price sits from it — the more honest version of the same idea.

Pi Cycle Top has failed the last two cycles in the same direction, for the same reason: both are calibrated on how steeply Bitcoin rose in its first decade. The 200-week moving average is the alternative trend line that adapts, because it is a trailing average rather than an extrapolation.