Pi Cycle Top

The Pi Cycle Top plots two simple moving averages of Bitcoin's daily close: the 111-day SMA and twice the 350-day SMA. The signal is a crossover. When the faster 111-day line moves up through the slower 350-day-×-2 line, the indicator declares a cycle-top warning. There is no second moving-average sequence, no oscillator, no threshold scaling — the entire indicator is the relative position of those two lines.

The two windows were chosen because their ratio approximates π: 350 ÷ 111 ≈ 3.153. Philip Swift documented the choice in his June 2019 piece The Golden Ratio Multiplier, where he wrote: “For the past three market cycles, when the 111DMA moves up and crosses the 350DMA × 2 we see that it coincides with the price of Bitcoin peaking” (Medium, 17 Jun 2019). In its tightest cases — December 2013 and December 2017 — the cross printed at or one day from the close-print top; other crosses led by months, as the history table below records. Today the gap between the two lines is −60.8%; spot is $65,837.03; the indicator reads far.

The input is the daily Bitcoin USD close history (see data sources). For every day t:

111DMA(t) = mean(price[t−110..t])
350DMA(t) = mean(price[t−349..t])
trigger(t) = 2 × 350DMA(t)

A signal is recorded on day t when 111DMA(t−1) ≤ trigger(t−1) and 111DMA(t) > trigger(t) — i.e. the faster line was at or below the slower-times-two line on the previous day, and is above it today. Crossings in the reverse direction are not signals; this is a one-directional indicator. Full derivation lives on the methodology page.

One choice is worth flagging. Some implementations use the daily-close ratio 111DMA ÷ 350DMA against a fixed threshold of 2.0 instead of the crossover. Mathematically equivalent at the trigger; the crossover form is the author-canonical one and the form the chart renders.

The single read on this chart is the gap between the two lines. The reading row above carries today's value as a percentage: when the 111-day MA is far below twice the 350-day MA, the cycle is in expansion; when it tightens to within a few percent, the indicator is on the watch list; when it crosses above, the historical pattern says the cycle is topping. The four regime labels Far / Diverging / Converging / Signal map to gap thresholds that capture the typical run-up to a cross. Because both lines are long moving averages, the gap moves on the order of weeks — Pi Cycle is a top-confirmation tool, not a same-day entry signal.

Walking the parallel arrays and emitting one row per published crossover surfaces the indicator's full history at a glance. For each cross we read the daily close on the cross date, locate the highest price in the 365 days that follow, and compute the drawdown from the cross over the next 200 days. The gap from cross to cycle peak is what makes Pi Cycle famous; the post-cross drawdown is what made it actionable. Both come straight from our daily-close history, with no smoothing.

The headline question for any current reader of Pi Cycle is the one the indicator itself does not answer cleanly: did the 2024 cycle peak come and go without a signal? The closest approach inside the run-up to the 14 March 2024 pre-halving high ($73,738 intraday) — defined as the maximum value of (111DMA − 2·350DMA) / (2·350DMA) observed in the window — is shown below alongside the realised price peak in that same window. The indicator never crossed.

Inside the early-2024 window, the closest the 111-day MA came to its 350-day-×-2 trigger was on 02 Jun 2024, when the gap was -26.4%. The realised cycle peak in that window was $73,097.77 on 14 Mar 2024. Bitcoin printed a new all-time high without the indicator firing.

The October 2025 ATH window tells the same story. Closest approach 19 Aug 2025 at gap -40.9%; realised peak $124,773.51 on 07 Oct 2025. Two ATH windows in a row without a Pi Cycle signal — the first time that has happened in our data.

Matt Crosby formalised the broader observation in late 2025: “the Pi Cycle Top Indicator failed to provide precise timing or price signals despite being closely watched by many traders” (Crosby, Dec 2025). The canonical reference page now carries the qualifier directly: the indicator “may cease to be relevant in this new market structure”. The 2024 and 2025 misses are the data behind that walk-back.

The 2021 cycle is the only one Pi Cycle has caught in part. The cross printed on 12 April 2021 at a daily close of $59,988 — two days ahead of the April local-peak close of $63,576 (which reached $64,863 intraday) — calling the top of that first run-up. Bitcoin then fell roughly 54% to a $30k summer trough before climbing back to a fresh ATH of $68,789 intraday in November 2021. The 111-day MA never re-crossed the 350-day-×-2 trigger on the second push higher. The indicator caught the first peak and missed the second.

The structural explanation is that 2021 was the first cycle with a bifurcated peak structure rather than a single blow-off. After the April top reset the moving averages, the 111-day MA could not re-accelerate fast enough to break above 2 · 350-day MA on the second leg — the second leg was higher in price but lower in momentum. That is the same reason the indicator has not fired on the 2024 or 2025 ATHs: this cycle is even less momentum-driven than the second half of 2021. Whether that pattern continues is the live question for any reader holding the chart open through 2026.

A slow second leg leaves the cross unprinted. Pi Cycle fires only when the 111-day MA accelerates hard enough to overtake twice the 350-day MA. When a cycle tops on lower momentum — a grind to a new high rather than a blow-off — the faster line never catches up and the signal simply never appears. That is exactly what happened to the November 2021 high of $68,789 intraday, where the second leg outran the first in price but not in pace, and what has happened across the March 2024 ($73,738 intraday) and October 2025 highs. Two consecutive ATH windows have come and gone without a cross — the first such stretch in the record. The canonical reference page now hosts a direct walk-back: the indicator “has worked during Bitcoin's adoption growth phase, the first 15 years or so of Bitcoin's life. With the launch of Bitcoin ETF's and Bitcoin's increased integration into the global financial system, this indicator may cease to be relevant in this new market structure” (reference).

The 111-day and 350-day windows are post-hoc fits. Swift selected them because they aligned with the 2013 and 2017 cycle peaks — not because of any structural derivation. The π-trivia is rounding: 350 ÷ 111 ≈ 3.153. With only four complete cycles to fit, two of those windows defining the signal, the danger is that the tuning describes the past more than it predicts the future. The recalibration risk is sharper here than for band models because the entire signal hangs on two narrow integer window sizes.

It is silent on cycle bottoms. Pi Cycle does not have a paired bottoming signal — the down-cross direction is not a published trigger. A reader using it for accumulation timing has to pair it with something else: the 200-week MA touchdown, MVRV-Z below 0.1, or the deep-capitulation band on the Rainbow. Pi Cycle answers the cycle-top question; bottom timing belongs to a different chart.

If you accumulate on a schedule, Pi Cycle is the wrong resolution to act on — but a fresh cross is a fair reason to slow buys, and the absence of a cross during a fresh all-time high (2024, 2025) is a reason to keep going rather than to assume the top is in. Treat it as a peripheral check, not a trigger.

If you are timing the cycle top, Pi Cycle earns its place precisely because it has called the exact close-print top day in three of four cycles. But the 2021 second-leg miss and the 2024–2025 blanks are the case for never reading it alone. Pair the cross with at least two of the 200-week MA, the Golden-Ratio Multiplier, and MVRV‑Z; when the cross fails, those valuation reads are what still flag the top.