Pi Cycle Top: the Bitcoin cycle-top indicator
Few indicators have such a striking record: the Pi Cycle Top has flagged Bitcoin's major tops within days. Understanding why it works —and why it might stop working— is understanding the cycles of this market.
Definition and calculation
The Pi Cycle Top, created by Philip Swift (Look Into Bitcoin), compares two long moving averages of Bitcoin's daily price:
Slow line = SMA 350 days × 2
The signal fires when the SMA111 crosses above the SMA350×2.
Where the name comes from
From a numerical curiosity with no real significance: 350 divided by 111 gives roughly 3.153, close to π. Worth saying early to strip away the mysticism: it is a catchy name, not a mathematical property of the indicator. The Pi Cycle stands or falls on its acceleration logic, not on a decimal coincidence.
The logic behind it
The two averages were not arbitrary in what they represent. The SMA111 reflects the average cost of buyers over the last three and a half months: what recent entrants paid, on average. The doubled SMA350 represents twice the average price of the last year.
For the first to reach the second, price must have multiplied relative to its annual average in very little time. That is not an arbitrary condition: it is the mathematical signature of sustained parabolic acceleration, exactly the behaviour of terminal euphoria phases, when price detaches from every historical reference.
The record
The cross has coincided, within days, with the euphoric tops of Bitcoin's cycles: the 2013 high, the December 2017 top and the April 2021 high. That retrospective precision made it one of the sector's most popular cycle indicators.
Three hits are impressive to look at and very little as evidence. What follows matters more than the record.
The limitations you need to know
Tiny sample
Bitcoin has had four complete cycles. Any pattern with n=4 can be statistical coincidence. Treating it as context is reasonable; treating it as law is not.
Overfitting risk
The numbers 111 and 350×2 did not come from a prior theory: they were chosen because they fitted the already known tops. That is fitting the model to the past, and it guarantees an impeccable record backwards while guaranteeing nothing at all forwards. It is the same problem any strategy optimized on historical data carries: the past can always be explained.
Changing market structure
The indicator needs parabolic phases to trigger, and that is its most serious weakness. Institutional flows —spot ETFs, corporate treasuries— tend to enter and exit more gradually than retail capital. If cycles lengthen or flatten, the cross could arrive late, early or not at all. That would not be a fault of the indicator: it describes a type of market that may be changing.
It only marks tops
It says nothing about bottoms or intermediate phases. For that there are complementary metrics: the Mayer Multiple (price divided by the SMA200, historical mean around 1.5), the AHR999 with its accumulation thresholds, or the Puell Multiple, which compares miner revenue with its annual average. All are explained in the halving cycle guide.
How to use it well
The Pi Cycle Top works as a context alarm, not as a tradable signal. And there is a nuance that gets overlooked: what is informative is not the cross itself, which arrives once almost everything has happened, but the speed at which the gap between the two averages narrows. When that gap closes fast, the market is in an acceleration phase and statistical caution rises week by week.
Combined with an extreme weekly RSI, falling Bitcoin dominance and hot valuation metrics, it sketches the classic late-cycle profile. No single one of those pieces justifies a decision.
Why it does not apply to other assets
It was designed and calibrated specifically for Bitcoin and its halving-linked cycles. Ethereum has no halving and no fixed issuance, and XRP has no issuance at all: applying an indicator calibrated on Bitcoin's supply structure to them has no basis. Their cycles, moreover, depend largely on Bitcoin itself and on internal capital rotations the ETH/BTC ratio describes far better.
Common mistakes
- Treating the cross as a sell order. It is a context alarm on one asset, not a tradable signal.
- Forgetting the parameters were chosen after the fact. Three hits with numbers fitted to the past are not three predictions.
- Applying it to altcoins, where it has no historical basis whatsoever.
- Expecting it to mark bottoms. It does not; other metrics exist for that.
- Watching only the cross and not the convergence. The useful information is in the speed of approach, not the instant of the cross.
Frequently asked questions
Has the Pi Cycle Top ever failed?
Its public record marks the 2013, 2017 and 2021 tops with day-level precision. But with only four cycles of history and parameters chosen after the fact, its future reliability is a hypothesis, not a fact. Market structure has also changed with ETFs and institutionalization.
Does the Pi Cycle Top work for altcoins?
It was designed and calibrated specifically for Bitcoin and its halving-linked cycles. Applying it to altcoins has no historical basis: their cycles largely depend on Bitcoin itself and on internal capital rotations.
Is there an equivalent indicator for market bottoms?
The author proposes the Pi Cycle Bottom, and metrics such as the AHR999 (below 0.45) or the Puell Multiple at its lows have historically coincided with cycle bottoms — with the same small-sample caveat.
Why is it called Pi Cycle?
Because of a numerical curiosity with no real mathematical significance: 350 divided by 111 gives roughly 3.153, close to π. It is a catchy name, not a property of the indicator. A decimal coincidence should not be mistaken for a theoretical justification; the Pi Cycle stands or falls on its acceleration logic, not on its name.
What exactly does it mean for the SMA111 to reach the doubled SMA350?
The SMA111 reflects the average cost of buyers over the last three and a half months; the doubled SMA350 represents twice the average price of the last year. For the first to reach the second, price must have multiplied relative to its annual average in very little time. It is not an arbitrary signal: it is the mathematical signature of sustained parabolic acceleration.
What is the overfitting risk with this indicator?
The numbers 111 and 350×2 did not come from a prior theory: they were chosen because they fitted the already known tops. That is fitting the model to the past, and it guarantees an impeccable record backwards while guaranteeing nothing forwards. It is the same problem any strategy optimized on historical data carries.
Can institutionalization invalidate the indicator?
That is the most plausible scenario. The Pi Cycle needs parabolic phases to trigger, and institutional flows tend to enter and exit more gradually than retail capital. If cycles lengthen or flatten, the cross could arrive late, early or not at all. That would not be a fault of the indicator: it describes a type of market that may be changing.
So how do you use it without getting it wrong?
As a context alarm, never as a sell signal. What is informative is not the cross itself but the speed at which the gap between the two averages narrows: when that gap closes fast, the market is accelerating. Combined with an extreme weekly RSI, falling dominance and hot valuation metrics, it sketches the late-cycle profile. No single piece justifies a decision.
Which other indicators should it be crossed with?
Those covering what it does not measure. The Mayer Multiple places price against its 200-day average; AHR999 and the Puell Multiple describe accumulation and capitulation zones; Bitcoin dominance indicates whether capital is rotating into speculative assets, which is typical of late phases. The Pi Cycle only speaks about tops, and only on Bitcoin.
Track this indicator in real time
Crypto Terminal computes it live alongside 17 other modules, with confluence across 6 timeframes.
Get it on Google PlayFree companion book: Technical Analysis of Crypto-Assets: Crypto Terminal · Edición en español