Cycle overlay
What a repeat of a past cycle would look like
Divide price by the trend and every cycle becomes the same kind of path, from a low near the floor to an overshoot and back. This chart replays each past recovery from this cycle’s low. The shape is the claim; the dollars come from the trend.
| Analog | Peak multiple | Valuation top | Dollar top | Implied price | × from today |
|---|
The valuation top and the dollar top are different moments. The trend keeps rising after the multiple rolls over, so a path's highest price can land months after its highest multiple: on the 2022 analog, ten months later and 11% higher. The implied price column reports the dollar top.
Are cycles speeding up? The timing record, and how the overlay is built
The overlay above replays each past cycle at its original speed. That is an assumption worth testing, because “cycles are getting faster” is the most common objection to it. Measured against the record, the answer splits in two.
| Cycle top | Bull leg low → top |
Halving → top | Top → top | Bear that followed top → low |
|---|
Tops and lows are the highest and lowest daily closes of each cycle; this cycle's low is the detected one and can still be undercut. Dec 2013 is shown for the record but excluded from the bull-leg and halving statistics: it predates the halving-driven regime and its 747-day advance is a clear outlier. Its bear is included in the bear baseline, which spans all three completed bears.
So the two halves of a cycle behave differently. The advance has been a metronome; the decline is what changed: this bear was both the shortest and by far the shallowest on record (−52%, against −77%, −83% and −85%). A shorter bear pulls the low forward without making the bull that follows any faster, which is exactly what the overlay assumes when it anchors on the low rather than on the calendar.
That gives two independent clocks for the next top, and they do not agree, because this cycle bottomed unusually early relative to its halving:
Four months apart, from two clocks that have each been accurate to within a few weeks across three cycles. That disagreement is the honest state of the question: not a forecast, and not a reason to prefer one number. Both are drawn from three completed cycles.
The clock that is speeding up
There is one sense in which cycles are unambiguously compressing, and it has nothing to do with behaviour. The power law's native clock is log time: the trend is a straight line only when age is plotted logarithmically. A cycle of fixed calendar length therefore occupies a shrinking slice of that clock as the network ages:
Those are the top-to-top spans of the last four cycles measured in log10 of Bitcoin's age. Each step takes roughly a quarter to a third less log-time than the one before it (−15%, −40%, −27%), in the only units the model actually uses. And at a constant 3.9-year calendar period the next ones fall to 0.077 by age 20 and 0.051 by age 31. This is arithmetic, not adoption: a fixed span divided by a growing age. It is also why calendar time, not log time, is the right axis for the overlay: across the modern cycles the calendar period varies by 1.9%, the log-time period by 42%.
The one thing the record actually says
Those are the peak trend multiples of the 2013, 2017, 2021 and 2024 tops. Each cycle overshoots less than the last: the diminishing amplitude the power law predicts as the network ages. So the four lines on the chart are not four equally likely futures. Reading the 2011 analog as a forecast is the least defensible thing you can do with this page; the most recent analog is the most relevant one, and even it may prove too generous.
Method & honesty
- Alignment. Past analogs are anchored at each bear's lowest daily close (2011-11-18, 2015-01-14, 2018-12-15, 2022-11-21). This cycle's anchor is the lowest trend multiple printed since its valuation top: provisional, and it can still be undercut. The other three anchor options let you assume the low is today, or still six to twelve months away.
- Absolute vs rebased. By default each analog replays its own historical multiples, so its peak is the valuation that cycle actually reached. Those analogs did not all bottom at the same multiple as we did (0.42×, 0.53×, 0.56×, 0.41×) so their lines start slightly off today's level. Rebase to anchor rescales each path to start exactly where we are, preserving the shape but shrinking or stretching the amplitude.
- The 2015 caveat. That bear's lowest price (Jan 2015) was not its lowest valuation: a deeper 0.365× came seven months later, in Aug 2015. Aligning on price lows is the standard convention, but it means the 2015 analog starts from a shallower valuation than the bear's true trough.
- Calendar time is not causality. The halving is the only exogenous clock Bitcoin has, and past lows sat at different distances from it. The 2028 halving marker on the chart shows where each analog's replay happens to fall relative to it, not a mechanism.
- n = 4. Four cycles is a small sample, and the trend line itself is fit to the same data. This is a descriptive overlay of what has happened, projected onto arithmetic: not a forecast, and not a distribution.
A replay of history against the trend, not a forecast. For where the power law puts the price on a given date, see the forecast; to plan on the floor instead of a repeat, plan your retirement on the floor.