Bitcoin Cycle Indicators
The classic top-and-bottom indicators, each recomputed nightly from one daily close series we keep ourselves (Bitstamp 2012 to 2017, Binance since) and stored as daily history you can download. Every one of them is a level, not a forecast: the page shows where each sits today, the full curve, and for the Pi Cycle the exact dates it crossed in the past. Puell uses our own Bitcoin node for miner revenue.
Pi Cycle crosses on record
Dates on which the 111-day average first rose above twice the 350-day average, from our own series. Past crosses coincided with cycle tops; the count is small and the next one need not behave the same way.
Pi Cycle Top
The 111-day moving average against twice the 350-day moving average. The cross of the fast line above the slow line has marked previous cycle tops within days.
Ratio at or above 1 = crossed. Below 1 = not crossed; the distance is how far the fast average sits under the slow one.
Mayer Multiple
Close divided by the 200-day moving average: how stretched price is against its medium-term trend.
Above 2.4 has historically marked overheated markets; below 0.8 deep value. Most days sit between 0.8 and 1.5.
200-week moving average
The 1,400-day average of daily closes, the slowest trend line in common use. Bear-market lows have tended to touch it.
Close below the 200W MA has been rare and short-lived; the multiple above it shows how extended a bull run is.
2-year MA multiplier
The 730-day moving average and five times that average, the two bands of the 2-year MA multiplier chart.
Below the 2-year MA has marked accumulation zones; above 5x the 2-year MA has marked cycle tops.
Golden ratio multiple
Close divided by the 350-day moving average. The golden ratio chart draws bands at 1.6, 2, 3 and 5 times that average.
Readings above 3 have been late-cycle; above 5 marked the 2013 and 2017 tops.
Profitable days
The share of all past days whose close was below today's close: how many past buyers are in profit.
100% means a new all-time high territory; values far below 100% show how deep a drawdown has cut into past buyers.
Stock-to-flow ratio
Coins mined so far divided by coins mined in the last 365 days, on the nominal issuance schedule. Doubles at each halving.
A supply statistic, published without any price model attached.
Puell Multiple
Daily issuance in USD divided by its 365-day average: miner revenue relative to its own yearly norm.
Above 4 has marked cycle tops; below 0.5 capitulation. History uses the nominal subsidy (our node is pruned); from 2026-09-02 the day's value uses the node's actual subsidy plus fees.
Bitfinex margin positioning
Aggregate margin long and short BTC positions reported by Bitfinex, recorded daily from 2026-09-02.
The long/short ratio is a positioning gauge for one venue, not the market.
No data recorded yet for this series.
Miner revenue (own node)
Subsidy plus fees per UTC day read from our own Bitcoin node with getblockstats, recorded from 2026-09-02.
Fees as a share of revenue show how much of miner income the mempool is paying for.
Anonymous readers see the last two years with zoom locked; members get the full series back to 2012 with zoom and pan.
Methodology and honest limits
Price series: daily closes from Bitstamp (2012-01-01 to 2017-08-16, backfilled once from the exchange's public OHLC endpoint) and Binance BTCUSDT (2017-08-17 to today, recorded nightly). Moving averages are simple averages of daily closes over 111, 200, 350, 730 and 1,400 days; a window that is not yet full produces no value rather than a partial one.
Stock-to-flow and the historical Puell Multiple use the nominal issuance schedule (144 blocks per day, halvings every 210,000 blocks). Real block production drifts a few percent from nominal and fees are excluded from history, because our node is pruned and cannot replay old blocks; from 2026-09-02 the nightly recorder substitutes the actual subsidy plus fees the node reports for that day.
Every value is written to the daily metric table and shipped in the CC0 metric-daily download, so the curves on this page can be reproduced and back-tested. None of these indicators is a signal: the site's engine does not trade on them and the copy does not predict from them.
FAQ
Has the Pi Cycle Top indicator worked every time?
It crossed near the 2013, 2017 and 2021 tops on our series, and the list above shows every cross we have on record. Three or four events are not a statistic; treat the indicator as a description of how far a run has stretched, not a timer.
Why do your numbers differ slightly from other sites?
Different price sources (some use a volume-weighted index, we use one exchange's daily close per era), different averaging conventions, and for stock-to-flow and Puell the nominal versus actual block schedule. The definitions above let you reconcile them.
Where is the history?
Every indicator is a metric in the free CC0 metric-daily dataset (CSV and JSON) with one row per day since the window first filled; the JSON API returns the same series.