chainmeter
EN | 中文 | ES

Tools · Historical

4-Year Cycle Top & Bottom Timing

Every past cycle top and bottom, detected by one fixed rule — a peak (or trough) confirmed only once price reverses 70% before making a new extreme. The same rule is then run forward from the next halving, mechanically, to show what a naive repeat of the historical rhythm would imply. This is jizhou-qiujian — marking the boat to find the sword — on purpose: a naive pattern extrapolation, clearly labeled as one, not a forecast.

Data: PriceUSD · Coin Metrics community API · pivot detection and projection computed live in your browser, nothing uploaded

Fetching data from Coin Metrics…

How a top or bottom is confirmed

The rule is the same one used everywhere on this chart, applied identically across all of history: track the running peak (or trough) price; once price reverses 70% away from it before setting a new extreme, that peak (or trough) is confirmed. The 70% threshold isn't tuned per cycle — it's a single fixed number, and it happens to be the threshold that reproduces the three cycle tops and bottoms most Bitcoin observers already recognize (Dec 2013, Dec 2017, Nov 2021 for tops; Jan 2015, Dec 2018, Nov 2022 for bottoms) without picking any dates by hand. Smaller thresholds pick up extra noise — the April 2013 mini-bubble, the March 2018 dead-cat bounce — that most people don't count as a full cycle.

What "mechanical extrapolation" means here

For the three completed cycles, this page measures how many days elapsed from each halving to its confirmed top, and from each halving to its confirmed bottom. It averages those two numbers, then simply adds the same offsets to the estimated date of the next halving. That's the entire method — no adjustment for shrinking volatility, diminishing returns, ETF flows, or anything else that has visibly changed between cycles. It is deliberately naive, which is the whole point of labeling it jizhou-qiujian: a fixed rule applied without regard for the fact that the boat (the market) has moved.

The current cycle, honestly

Loading…

n = 3. Every average on this page is computed from exactly three completed halving cycles. That is not a statistically meaningful sample by any normal standard — it's the entire available history. The "projected window" bands on the chart are the min-to-max range across those three cycles, not a confidence interval, and there is no reason a fourth or fifth cycle has to fall inside it.

The three completed cycles happened in three different markets: pre-institutional retail speculation (2013), first mainstream retail wave (2017), and a DeFi/COVID-liquidity/leverage cycle (2021). The current and future cycles are unfolding in an ETF-and-treasury-company market that didn't exist before. A pattern built entirely from pre-ETF cycles projecting into a post-ETF world is exactly the kind of assumption "marking the boat" gets wrong.

The next halving date itself is an estimate (constant 10-minute average block time from block 840,000), not a fact — see the Halving Countdown page for how that estimate can drift.

FAQ

How is a "cycle top" defined here, exactly?
A confirmed top is a local price peak that was followed by at least a 70% drawdown before the next all-time high was set. The 70% threshold is fixed and applied identically to all of history — it is not tuned per cycle. It happens to be the threshold that cleanly reproduces the three widely-recognized cycle tops (Dec 2013, Dec 2017, Nov 2021) with no manual date-picking.
Why call this "jizhou-qiujian" (marking the boat to find the sword)?
Because that is exactly the method: take the average number of days from halving to top/bottom across three past cycles, and add that same offset to the next halving, mechanically, regardless of whether market conditions have changed. It is a naive extrapolation, not a model — the name is meant as an honest warning, not a strategy.
Has the current cycle's top already happened?
Unconfirmed. Price set a local high that fits the historical timing window, but under the same 70% rule applied to prior cycles, that peak isn't confirmed as "the" cycle top until price has drawn down 70% from it without setting a new high first. Check the stat cards above for the live drawdown-so-far from that candidate peak.
Why is the projected window so wide?
Because the three historical data points disagree with each other by a lot — the fastest cycle topped 371 days after halving, the slowest 546 days, a spread of nearly six months. Showing a single date would manufacture false precision the underlying data doesn't support.