9 min read

In 2022, a Bitcoin cycle bottom fell below the previous peak for the first time. In 2026, it's happening again.

Five peak-to-peak epochs since 2011, one halving in each. Both the upside multiple between peaks and the crash ratio to the following bottom compress cycle after cycle. In 2021 the order of those two ratios flipped for the first time. The one number that has not moved is the twelve-to-thirteen-month peak-to-bottom leg.

2013201720212025

The Bitcoin asset page carries a price. This page describes the structure that price sits inside. Nothing here is a forecast; the ratios are read directly out of the peak, bottom and halving dates of every cycle to date, and the finding is that the numbers are getting smaller as the asset matures.

Sixteen years, on a log scale

Monthly closes from $0.07 (Jul 2010) through the current print. Triangles mark each cycle peak (up) and the bottom that followed (down), coloured by epoch.

Log scale · monthly closes · Coin Metrics PriceUSD series through May 2026, live price feeds thereafter.

The five epochs

Each cycle runs peak-to-peak with exactly one halving in between. The peak lands twelve to eighteen months after the halving that precedes it — a window that has stabilised at seventeen to eighteen months across the last three cycles.

CyclePeak datePeak (USD)Halving Halving → peakBottomDrawdown
20112011-06-08~$31——~$2−94%
20132013-11-30~$1,1562012-11-2812.1 mo~$172−85%
20172017-12-17~$19,7832016-07-0917.3 mo~$3,189−84%
20212021-11-10~$69,0442020-05-1118.0 mo~$15,598−77%
20252025-10-06~$126,1982024-04-2017.5 mopendingpending

Both slow drifts continue. Halving-to-peak stabilised at seventeen to eighteen months. The drawdown depth column keeps shrinking: −94, −85, −84, −77 — and the 2025 cycle is a work in progress.

Drawdown depth from each cycle peak to its bottom. The teal 2025 est. bar (−74%) continues the shrinking trend to a confirmed bottom — the −52% is the current incomplete low from the June 2026 print, not the floor.

Each Bitcoin cycle gains less, and falls less

Cycle to cycle, the upside multiple between peaks shrinks. The crash ratio down to each bottom shrinks with it. Bitcoin is maturing on both sides of the trade — smaller gains and smaller collapses.

CyclePeakBottomPeak ÷ prev peakPeak ÷ own bottomBottom ÷ prev peak
2013$1,156$17237.3×6.72×5.55×
2017$19,783$3,18917.1×6.20×2.76×
2021$69,044$15,5983.49×4.43×0.79×
2025 (low so far)$126,198$58k1.83×2.18×0.84×
2025 (trend est.)$126,198~$33k1.83×3.85×0.48×

The 0.84× reading for 2025 breaks the declining sequence because the low isn't in — following the drawdown trend (−85% → −84% → −77% → ~−74%) puts a confirmed bottom near $33k and the ratio at 0.48×, which continues the pattern: 5.55 → 2.76 → 0.79 → 0.48.

Log scale. Both ratios fall in lockstep; in 2021 the crash ratio (orange) overtakes the peak-to-peak gain (blue) for the first time.

the 2021 flip The crash now outweighs the climb

In 2013 and 2017 the upside multiple was the bigger number — each top dwarfed the last (37×, 17×) far more than it later fell (6.7×, 6.2×). In 2021 the order reversed: peak÷prev (3.49×) dropped below peak÷bottom (4.43×), and 2025 continues it. That reversal is the same fact as the cycle bottom falling below the previous cycle's peak — the right-hand column crossing 1.0×.

Both cycles show the same structural arc, measured from each cycle's peak. Only peak-to-first-breach compresses; the deep low, bounce peak and confirmed bottom all land at nearly matching months-from-peak.

Event2021 cycle · Peak Nov 20212025 cycle · Peak Oct 2025
Peak → first breach 7 mo · Jun 2022 below $19.8k 4 mo · Feb 2026 below $69k
Peak → deep low after breach 7 mo · $17.6k intra 8 mo · $58k intra
Peak → bounce peak 8 mo · $23.4k (Jul 2022) ~11 mo · $86k current (Sep 2026)
Peak → confirmed bottom 12 mo · $15.6k (Nov 2022) 12–13 mo projected · Oct/Nov 2026

Peak-to-bottom across three prior cycles ran 13.5 → 11.9 → 12.4 months — a spread of only 1.6 months, the article's "one constant." For $58k to be THE 2025 bottom, peak-to-bottom would compress by ~4 months in one cycle, breaking that spread by more than the entire cycle-to-cycle variation of the prior three combined.

A reader hitting this page in the second half of 2026 is inside the mid-cycle bounce phase. The current +48% off the June $58k low is proportionally larger than 2021's +33% Jul–Oct 2022 bounce off $17.6k, but at the same phase of the arc — before the confirmed bottom.

The 2017 cycle warns how misleading a mid-cycle bounce can be. Every prior cycle followed a first-low → bounce → next-low-attempt shape; only 2021's next-low went lower:

— 2013: Jan 2015 $217 first low → Jul $285 (+31%) → Aug $231 retest (held 6% above).
— 2017: Dec 2018 $3.19k first low → Jun 2019 $10.8k (+240%) → nine-month decline to March 2020 COVID $3.85k intra (held 22% above).
— 2021: Jun 2022 $17.6k first breach → Jul-Oct $23.4k (+33%) → Nov re-break to $15.6k (the only prior cycle to make a new lower low after the bounce).

Whether $58k is an unprecedented early bottom (2013/2017 pattern of a first low that holds) or the first of two with a Nov 2026 re-break still ahead (2021 pattern), stays an open question — if the pattern holds, it resolves within weeks.

2013 bot vs '11 peak
5.55× ↑
2017 bot vs '13 peak
2.76× ↑
2021 bot vs '17 peak
0.79× ↓
2025 low vs '21 peak
0.84× now
~0.48× est. ↓

A buyer of the 2017 top was underwater at the 2021 bottom; a buyer of the 2021 top is underwater at the 2025 low. In the first two cycles, that never happened — every floor sat above the prior ceiling.

The breach depth is an open number. In 2021, the low landed 21% below the 2017 peak ($15.6k vs $19.8k). In 2025 so far, the low is 16% below the 2021 peak ($58k vs $69k) — but the cycle bottom is not in, and every prior mature cycle was still falling at this point. "The breach depth is also shrinking" would fit the rest of this page neatly, but it stays a hypothesis until a confirmed floor arrives. Three brackets if the downtrend has more to run: hold at ~16% and the current low is close to the floor; match 2021's 21% and the bottom sits near ~$54.5k; deepen against the shrinking trend to 25% and the floor is closer to ~$52k.

Where the monthly crosses actually fire

The same price series with the two moving-average systems overlaid. Red triangles are death crosses (fast below slow), teal are golden crosses (fast above slow). These land late on the price — they are confirmations, not calls.

Price 10-mo SMA 20-mo SMA death cross golden cross
Canonical simple-moving-average cross — 10-month vs 20-month SMA. Log scale, monthly.
Price 10-mo EMA 21-mo EMA death cross golden cross
EMA cross — 10-month vs 21-month EMA. Faster to weight recent price, but still lagging. Log scale, monthly.

Four checkpoints, and the time between them

Each cycle has four measurable points: the cycle peak, the death cross (monthly 10-EMA below 21-EMA), the bottom, and the golden cross. The intuitive order — peak → death → bottom → golden — held only in 2021. In 2013 and 2017 the monthly EMA is so slow that the bottom arrived roughly two and a half months before the death cross. The monthly death cross is a trailing confirmation of the decline, not an early warning; the golden cross is the reliable structural signal for the new uptrend.

CyclePeakDeath crossBottomGoldenPk→DeathDeath→BotBot→GoldPk→Bot
20132013-112015-032015-012016-0116.0−2.512.513.5
20172017-122019-022018-122019-0514.4−2.55.511.9
20212021-112022-082022-112023-119.7+2.712.312.4
20252025-102026-06pendingpending~8———
Peak→Bottom barely moves (~12 months); Peak→Death cross keeps shrinking as bears confirm faster.
~12 mo

Peak → bottom is the constant

13.5, 11.9, 12.4 months across three mature cycles — even as drawdown depth shrank from −85% to −77%.

5–12 mo

Bottom comes before the golden cross

Every cycle, the low landed months ahead of the golden cross — a lagging confirmation.

order flips

The death cross lags too

Peak → Death → Bottom → Golden held only in 2021. In 2013 and 2017 the bottom came ~2.5 months before the death cross.

16 → 8 mo

Bears confirm faster

Peak → Death cross shrank 16 → 14.4 → 9.7 → ~8 months — the maturing-asset signature.

Where the 2025 cycle stands today

The 2025 peak was $126,198 on 6 October 2025, seventeen and a half months after the April 2024 halving. Bitcoin traded into the $58k–$62k range in late June 2026 (a drawdown of roughly −52% from the peak), printed a first monthly death cross on the 10/20 SMA in May and on the 10/21 EMA in June, then recovered to trade in the mid-seventies through September 2026. The current print is $84,570, a drawdown of about −33% from the peak — above the June low but still well within the cycle's downtrend envelope.

Peak
$126,198
2025-10-06
June low
$58–62k
−52% from peak
Now
$84,570
−33% from peak
Death cross
May–Jun '26
~8 mo after peak

The June ~$60k print sits at about eight months from the peak. Every prior mature cycle was still falling at that point — the historical peak-to-bottom leg is twelve to thirteen and a half months. Two readings are both live and neither is proved yet:

Both remain consistent with the data at hand. The site does not adjudicate between them — a monthly golden cross (10 back above 21) is the reliable structural confirmation, and by the historical pattern it trails the eventual bottom by five to twelve months. It has not fired yet.

Where it could bottom, if the pattern holds

Continuing the shrinking-drawdown trend (−85%, −84%, −77%, then something shallower) gives a 2025-cycle bottom in the −64% to −74% range: approximately $33k to $45k. The already-shallower drawdown to date hints at a shallower still floor near $50–60k if the compression accelerates.

Scenario bottoms vs the $126k peak (grey) and current $84,570 (blue). The teal floor (~$30k) and the blue current line bracket the primary pattern-continuation band ($30–60k); the red −82% line ($22.5k) is the deeper average tail. Linear scale, $ thousands.
$16–27k

Deep / shock · −79% to −88%

Ratio 0.23–0.39. The raw bottom-÷-prev-peak extrapolation, a COVID-style cross-market shock, and the simple −82% historical-drawdown average all land here. Only in play if the shrinking-drawdown pattern breaks.

$30–36k

Bearish / trend holds · −71% to −76%

Ratio 0.43–0.52. The drawdown-percent method: if crashes keep shrinking only at the gentle historical pace, 2025 still mirrors a classic deep cyclical bear.

$40–46k

Blended / moderate · −64% to −68%

Ratio ~0.58–0.67. Partial maturation: drawdown shallower than every prior cycle, but the current low isn't quite it.

$48–60k

Moderate / maturation · −52% to −62%

Ratio 0.70–0.87. ETF/institutional bid holds, the −52% already printed is near the floor, and the current cycle bottoms shallower than any before.

The bearish cases are kept in because exogenous shocks override internal patterns. March 2020 cut BTC by half in a day and WTI crude futures went negative; a structural institutional bid does not floor price in a shock, it just means patient capital waiting for a lower price.

The bear-market timing itself

A cycle's bear has two phases: the decline (peak → bottom) and the accumulation window (bottom → the golden cross that confirms the next uptrend). The decline is remarkably steady; the accumulation window is where patient capital builds positions before the markup, and it has run anywhere from ~5.5 to ~12.5 months.

Decline (peak → bottom)Accumulation (bottom → golden cross)
Each bar is the full time underwater, split into the decline and the accumulation window. Months.

Mapping onto the current cycle: the October 2025 peak plus a ~13-month decline points to a bottom around Q4 2026, and a further ~6–12-month accumulation window puts the next confirmed uptrend somewhere in mid-to-late 2027. The scenario prices above are an accumulation range over that window, not a single-day level.

The one constant worth trusting

Every other number in the dataset moves. Peak-to-peak multiples move. Peak-to-bottom ratios move. Recovery multiples collapse cycle after cycle. Even the peak-to-death-cross delay is shrinking. But the peak-to-bottom leg does not move: 13.5, 11.9, 12.4 months. If that constant holds a fourth time, the 2025 cycle bottoms around October or November 2026. Price is the noisy variable in Bitcoin cycles; the duration of the fall is the stable one.

Data and method

Daily reference price from the Coin Metrics community dataset (PriceUSD), July 2010 through May 2026, plus live price feeds thereafter. Monthly closes; EMA(10), EMA(21), SMA(10), SMA(20); cross defined as the sign change of (fast − slow). Peaks and bottoms are cycle extremes; timings are day differences ÷ 30.44. The 2011 cycle is excluded from the cross tables — at sub-$30 prices the monthly EMAs never produced a clean isolated cross. Monthly MA crosses are inherently lagging; everything here is descriptive of past structure, not a trading signal or a forecast.

Related

Bitcoin on CoinSpectrum — the live asset page: price, market cap, consensus family, circulating supply, and every wrapped or staked claim on BTC the site tracks.

The proof-of-work hashrate leaderboard doesn't exist — why the security number every other tracker shows is a category error; the algorithm-first view of Bitcoin's mining base.

Why we count BTC, BCH and BSV as three cryptocurrencies — Bitcoin's cycle structure is unique to Bitcoin the chain, not to everything that shares the ticker.

Every proof-of-work cryptocurrency — Bitcoin ranked among the family it defined.