We can't call the bottom. So here's an honest recovery band instead.
Bitcoin sits about 50% below its October-2025 high of ~$125K, roughly eight months into the drawdown. Nobody can tell you when it bottoms or where it lands — so we draw the opposite of a target: a band. Standing on today's state — this deep below the high, this many months in — we project a floor, a middle, and a ceiling 12 to 24 months forward, built so about 8 of every 10 past recoveries from a state like this would land inside it.
The band is wide on purpose. Past recoveries diverged so much that an honest interval has to span a deep-bear floor and a new-high ceiling at once. The work in the parts that follow is showing exactly what that width is built from — and being honest about which edge today's evidence leans toward.
recovery_cone
Line up every past bear by months from the high. Today's path stands out.
The four-year halving clock can't tell you how far you've fallen — so we anchor on the high instead. Each line below is one completed Bitcoin bear: how far below its all-time high the price sank (down the chart) as the months ticked past since that high (across the chart). Today's drawdown from the October-2025 high is drawn in blue, with a marker at where we stand now.
The shape is consistent across bears, even though their sizes were not: a long slide to a bottom around month 12–13, then a multi-year climb back. The catch is where today sits on that shape.
bear_paths
The deeper the drawdown, the bigger the recovery that has followed.
Part 2 is the near-term caution. This is the other side of the same history — and it is genuinely encouraging. Sort Bitcoin's past by how deep below the high it was and how long since that high, then ask what the next two years looked like from each state. The pattern is monotone: the deeper the hole, the larger the climb that followed.
We show the modern era (2018 on) so the numbers are repeatable — the 2011–2013 hyper-growth years produced 1,000%-plus moves that no longer apply, and including them would only flatter the case.
forward_return_heatmap
phase_returns
A real edge a year out. None two years out. We tested it the honest way.
A gradient that looks clean in hindsight can still be useless out of sample. So we ran the strictest test the tiny sample allows: stand at each past bear's halfway point — about 50% below the high, where Bitcoin is today — and guess its next year and two years using only the other bears as a guide. Then measure how far off that guess was, versus the lazy guess that the price simply stays put.
This is the leave-one-bear-out test, and it splits cleanly by horizon.
recovery_skill
An honest clock: where you are in the bear, and what that has meant — both sides.
This is not advice to buy or sell, and it is not a price target. It's an honest reading of where Bitcoin sits in its drawdown — calibrated, balanced on both tails, and explicit about exactly how much of the pattern survives a stress test (a one-year direction) and how much doesn't (a two-year size).
From ~$62K today (~50% below the high, ~8 months in), a 1-year recovery band of about $36K–$170K (middle ~$83K) and a 2-year band of ~$48K–$422K (middle ~$142K). The old high near $125K sits only at the optimistic edge — past full recoveries took 28–39 months. A calibrated range, never a target.
Every past Bitcoin bear was deeper than today's at the same point, and each kept falling another 4–5 months before bottoming around month 12–13. Shallower and earlier is not 'the worst is over' — it's why the near-term read leans to more downside first.
The deeper the drawdown, the bigger the two-year recovery that has followed — deep states typically ran ~2.3–2.9× versus sideways-to-lower near a high. That upside gradient is real and balanced against the near-term caution; but the spread inside each state is enormous, so it tilts the odds rather than timing the move.
The leave-one-bear-out test gives a one-year directional edge (the bear usually isn't over) but no two-year magnitude skill — past recoveries diverged too much for one bear to predict another's size. Trust the direction at a year and the band's width at two; don't trust the exact middle.
- —The recovery band's width is the trustworthy part; the middle line is only weakly informative, and a year out the evidence leans toward the lower half first. Read the floor and the spread, not a target.
- —The one-year directional edge is real but narrow, and the two-year magnitude has essentially no out-of-sample skill — recoveries diverged from 0.9× to 2× from the same depth. The cone is wide on purpose.
- —Every number here rests on about three completed bears. The 'you are here' alignment is loose because each bear hit this depth at a different month (0.5, 1.5 and 6 months in), so even the near-term read is suggestive, not proven.
- —The upside gradient is shown for the modern era (2018 on). The full-history version looks even stronger, but it leans on Bitcoin's early hyper-growth years, which are not repeatable — so we publish the more conservative modern band.
- —Descriptive research about uncertainty, not a price prediction or advice. Bitcoin can lose most of its value in a year — as it has four times in this dataset — and a pattern across three bears can break on the fourth.
Descriptive research, not financial advice. Bitcoin is volatile and can lose most of its value in a year — as it has four times in this dataset. A pattern that has held across three bears is an assumption about the next one, not a fact.
Appendix — all 5 charts & the method →