Bitcoin's blow-off regime is thinning.
Count the days each cycle spends in the euphoric top band of the Rainbow regression and a quiet structural change appears: the manias are getting shorter. Not smaller in dollars — shorter in time.
As of 12 Jun 2026Under the live fit, the euphoric band absorbed 26.3% of all daily closes in its busiest closed cycle, then 12.9% of the most recent one — and 0.0% of the current cycle so far. The fraction of Bitcoin's life spent in mania has collapsed from its early-cycle peak and never recovered.
Counting days, not peaks
The companion note shows that each cycle's single most extreme day has been less extreme than the last. This note asks a different question: how much time does each cycle spend in the euphoric regime? Every daily close is assigned a Rainbow band under the live fit; the table counts the days landing in band 9 — the open-ended top band where the 2013 and 2017 blow-offs lived — within each halving-to-halving window.
| Cycle | Days | Euphoric days | Share |
|---|---|---|---|
| Cycle 1 (2010 → 2012) | 864 | 127 | 14.7% |
| Cycle 2 (2012 → 2016) | 1,317 | 347 | 26.3% |
| Cycle 3 (2016 → 2020) | 1,402 | 165 | 11.8% |
| Cycle 4 (2020 → 2024) | 1,439 | 185 | 12.9% |
| Cycle 5 (2024 → today) | 785 | 0 | 0.0% |
What the thinning means
Early Bitcoin spent whole seasons in mania — long stretches where price sat far above any defensible trend estimate. No cycle since has spent half as much of its life there — Cycle 4 did edge back above Cycle 3, so the decline is not perfectly monotone — and the current cycle has yet to print a single euphoric day. The market still overshoots, but the overshoots resolve faster and reach less far. That is what maturation looks like in residency terms: the volatility series tells the same story from a different angle.
For anyone holding through cycles, the implication is the same as the companion note's, with a time dimension added: not only should you not expect the next top to reach as far above trend as 2017 did — you should not expect the window to linger. The 2013 and 2017 cycles gave sellers weeks inside the top band. Recent cycles have given days, or none.
Where this argument fails
Half lens, half market. The regression flattens as new data arrives, and a flatter fit re-scores past tops downward and makes today's band 9 harder to reach. Part of the decline is therefore mechanical. It is not all mechanical — the ordering survives on fixed historical fits — but the magnitudes are fit-dependent.
An open cycle is an unfinished sentence. Cycle 5's zero is provisional: one future blow-off would rewrite its row, and with only three closed cycles the trend is small-n by construction. The honest claim is a consistent direction, not a guarantee.