Bitcoin Cycle Monitor

Computing wavelet transform… this may take a moment on first load.
Overall Market Activity

A single percentile ranking Bitcoin's current rhythmic energy against every 30-day window since 2010. Backtesting against the full price history shows a meaningful signal at the High level (70–85th percentile): those periods have produced a median +50% price return over the following 90 days, with a 75% win rate. Extreme readings (>85th percentile) have accompanied Bitcoin's strongest bull runs — but occur on ~13% of all trading days (more often than High) and show weaker 180-day returns than High, indicating a late-cycle character. Below the 70th percentile, the score identifies the structural character of the market rather than predicting returns.

Percentile
Rhythm bands

Which timescales are shaping Bitcoin right now? And how many are firing simultaneously? To answer this, Bitcoin's price history is decomposed into six rhythm bands: from short weekly swings to multi-year macro cycles. Historical average shows each band's long-run dominance relative to the historically strongest band (= 100). Activity shows where the current energy level ranks within that band's own history. A band is Active when its 30-day energy is above the hysteresis zone. The base rate shown per band (e.g. "active 58% of history") indicates how frequently that band activates historically — a higher base rate means each individual activation carries less structural information.

Cyan line = this band's historical mean (reference only).   Yellow markers = hysteresis zone (45th–55th pct): band activates at 55th, deactivates at 45th — state is preserved inside this zone.
Bitcoin Scaling Law — Power Law & Log-Periodicity

Bitcoin's price follows a Power Law in log-time with exponent k ≈ 5.68 (R² ≈ 0.96 since 2016). Around this structural trend, price oscillates log-periodically with fundamental frequency ω ≈ 9.01 — each cycle compresses by factor λ ≈ 2.01 in log-time (approximately an age-doubling). These two scale-invariances are not independent: they are coupled through the Hidden Coupling Constant C = k·ln(λ) ≈ 3.96 — the same renormalisation-group structure observed in critical phenomena in statistical physics. Each completed major cycle raises the structural price floor by exp(C) ≈ 52×. The model projects the next fundamental peak near mid-2028 (Bitcoin age ≈ 19.5 years).

5.68
k — Power Law Index
9.01
ω — Log-Periodicity (fund.)
2.01
λ — Cycle Compression
≈ 3.96
C = k·ln(λ) — Hidden Coupling
Current Price (USD)
Deviation from Power Law
Computing model…

Bitcoin Cycle Monitor — Explanation, Benefits, and Methodology

Most market tools tell you where Bitcoin's price is. The Bitcoin Cycle Monitor tells you what the market's rhythmic structure looks like beneath it — which cycles are active, how unusual the current energy level is by historical standards, and what comparable periods have produced in the past.

What is the Bitcoin Cycle Monitor?

Bitcoin does not move randomly. Beneath its daily volatility, recurring rhythmic patterns — from multi-week swings to multi-year macro cycles — have consistently shaped price behavior since 2010. The Bitcoin Cycle Monitor uses a Continuous Wavelet Transform (CWT) to decompose Bitcoin's log-returns into six distinct frequency bands, measures the energy each band currently carries, and summarises that into two outputs:

  • Overall Market Activity score — a single percentile that ranks today's combined cycle energy against every comparable 30-day window since 2010. A high score means the market is unusually rhythmically active by historical standards.
  • Cycle Equalizer — a band-by-band view showing which timescales are currently engaged, how long they have typically persisted once active, and — where a statistically confirmed cycle exists — an estimate of the current phase and days to the next trough.

A third chart — the Bitcoin Scaling Law by Stephen Perrenod & Giovanni Santostasi — provides the long-run structural backdrop for both. It shows where the current price sits relative to Bitcoin's Power Law trend and its log-periodic cycle structure, and projects the composite model forward to 2045. Where the Wavelet tools answer "what is the market doing right now?", the Scaling Law answers "where are we in the multi-year structure?"

This is a context tool, not a timing tool. Individual band activations lag the price move significantly: shorter bands (Short-term, Mid-term, Quarter) typically activate 40–61% above the prior 6-month trough, while longer bands (Annual, Halving) activate 107–117% above it. The monitor is useful for confirming a structural trend, not for identifying its start.

Note on terminology: period is the length of each oscillation a band captures. Run length is how long that band's energy stays elevated above its historical median — a completely separate quantity. A band capturing 4-day oscillations can remain structurally active for months, just as a musical note at a fixed pitch can be held for any duration.

Why Is the Bitcoin Cycle Monitor Important?

Standard price charts show momentum and trend, but cannot distinguish between a quiet drift and a structurally driven move. The Cycle Monitor fills that gap by quantifying the rhythmic energy underneath price action.

The most important finding from backtesting the full price history since 2010: periods when the Overall Market Activity score enters the High range (70–85th percentile) have produced a median +50% price return over the following 90 days, with a 75% win rate — the strongest and most consistent threshold in the data. Knowing that the market is structurally in or approaching this zone changes the weight a long-term investor places on other signals.

The Extreme range (>85th percentile) has accompanied Bitcoin's strongest bull runs, but requires careful reading. First, the label is misleading: Extreme has historically occurred on 12.9% of all trading days — more frequently than High (8.3%). It is not a rare spike; it describes a sustained high-energy phase that can last for months. Second, while the 90-day median is +89% — higher than High — the 180-day median drops to +162%, well below High's +241%. The market captures much of the upside quickly, then gives back relative performance. This is the empirical definition of late-cycle: strong short-term momentum, weaker long-term continuation. Both sides matter; neither should be ignored. ⚠ Fewer than 10 independent observations — these figures carry wide uncertainty.

Below the 70th percentile the score identifies structural character rather than predicting returns. Counterintuitively, 0 active bands has historically produced the highest median 90-day return (+22.7%) — likely because quiet, low-energy phases often precede the next structural build-up. A low score does not mean poor forward returns; it means the current market lacks the rhythmic energy that would confirm a structural trend is already underway.

Why is this relationship non-monotone? Band activations lag the price move (shorter bands typically activate 40–61% above the prior trough). By the time multiple bands are simultaneously active, a substantial portion of the cycle's upside has often already been captured. More active bands therefore reflect a market that is further into a structural move, not one that is just beginning. The absence of active bands reflects a market in reset — historically a better entry context than a fully-loaded cycle. This is why the Equalizer should be read as a structural description, not a return forecast.

Who Can Benefit from the Bitcoin Cycle Monitor?

  • Long-term investors who want to know whether current cycle energy is historically elevated or quiet before sizing a position — particularly whether the activity score is approaching or within the High range.
  • Active traders seeking a structural backdrop: the Cycle Equalizer shows which timescales are engaged and how long active bands have typically persisted, helping calibrate trade horizon and avoid entering late in an extended structural move.
  • Risk managers who want an objective, data-driven read on whether the market is in a structurally active phase or a quiet, low-energy period — and what each regime has historically implied for drawdown risk.
  • Analysts and researchers studying Bitcoin's multi-scale cyclical structure and the empirical relationship between wavelet energy and forward price behaviour.

Practical Applications of the Bitcoin Cycle Monitor

The monitor is most useful as one layer in a multi-signal framework. Concrete ways to use it:

  • Confirming a structural trend: When the Overall Market Activity score enters the High range (70–85th percentile) and the Quarter or Annual band activates, the market is likely mid-move in an extended structural trend — not at its start, but not necessarily near its end either.
  • Recognising late-cycle conditions: An Extreme reading (>85th percentile) alongside only short-term and mid-term bands active suggests fast momentum without structural depth — a pattern that has historically preceded corrections at the 180-day horizon more often than High readings do.
  • Watching for structural deterioration: When the number of simultaneously active bands declines, the historical data suggests more caution than when bands are building. A drop to 4 active bands has produced a median 90-day return of −30.0%; a drop to 3 bands −4.8% (both with fewer than 10 independent observations — wide uncertainty). The direction of change in the Equalizer carries information beyond the current count: a falling count often matters more than the count itself.
  • Reading a quiet market correctly: When 0 bands are active and the score is low, the monitor is not flashing a warning — it is describing a structurally quiet accumulation-type phase. The highest historical median 90-day return (+22.7%) belongs to exactly this condition. Quiet is not bearish.
  • Understanding what "Short-term only" means: When the Short-term band is the only active band, the situation is different from Short-term being active alongside longer bands. In isolation — all five longer bands inactive — this pattern has occurred on ~187 of 5,700+ trading days (3.3% of history) and produced a median 90-day return of +7.2%, below the baseline of +14.7% for all other conditions (8 independent observations). Short-term energy alone, without structural support from longer cycles, does not confirm a trend.
  • Combining with other indicators: The Scaling Law chart on this page shows where the current price sits relative to Bitcoin's long-run Power Law trend and its log-periodic cycle structure. Cross-referencing the Wavelet activity score with the Scaling Law deviation gives a two-dimensional read: high wavelet energy near the upper log-periodic band signals a structurally late and rhythmically active market — historically the condition most associated with sharp corrections over the following 180 days. Low wavelet energy with the price near or below the Power Law trend describes the opposite: structural reset with limited near-term rhythmic momentum, but historically the strongest entry context for long horizons.

How the Bitcoin Cycle Monitor is Calculated (Methodology)

Bitcoin's daily log-returns since July 2010 are processed through a Morlet Continuous Wavelet Transform (CWT) — an FFT-based implementation (O(N·S·log N)) that decomposes the signal into a time-frequency power spectrum, revealing how much energy each rhythm band carries at every point in time.

Overall Market Activity score: For each of the six bands, the 30-day average energy is ranked as a percentile against that band's own full history (static reference back to 2010). These six percentiles are combined into a single weighted score — longer cycles receive higher weight (weights 0.5 to 3.0, linear spacing) because structural macro cycles carry more informational value than short-term noise. The resulting score is itself ranked against all historical days to produce the final 0–100 percentile.

Cycle Equalizer — Active/Inactive status: Each band's 30-day energy is compared to its own historical distribution using a hysteresis rule: a band activates when its percentile reaches the 55th percentile and deactivates when it falls back to the 45th percentile. The dead zone between 45 and 55 preserves the previous state — this prevents flickering around the threshold. The yellow markers in the Activity bar visualise this zone directly.

Cycle phase and days to trough: Where a statistically confirmed cycle exists within an active band — validated via FFT surrogate testing (200 phase-randomised surrogates, Theiler 1992) — the monitor estimates the current phase using a Hilbert transform on the bandpass-filtered signal, and derives days to the next trough from the dominant tactical period. These estimates are only shown when the wavelet ridge has been stable for at least 90 consecutive days and multi-cycle coherence is at least 50%. Cycles that do not meet these thresholds are marked "unconfirmed."

Bitcoin Scaling Law — Power Law & Log-Periodicity

The Scaling Law chart rests on a different mathematical foundation — one derived from statistical physics rather than signal processing. The core observation, documented by Giovanni Santostasi and Stephen Perrenod, is that Bitcoin's price history exhibits two simultaneous forms of scale invariance that are not independent of each other.

Power Law (continuous scale invariance): In log-log space, Bitcoin's price follows a straight line against time since the Genesis Block (3 January 2009). An OLS regression of log(price) on log(age in days) produces a slope of k ≈ 5.68 with R² ≈ 0.96 — meaning the Power Law alone explains 96% of the variance in log-price over Bitcoin's full history. The intercept is fit directly from the data on each daily update.

Log-periodicity (discrete scale invariance): The residuals around the Power Law are not random. They oscillate with a frequency ω ≈ 9.01 in log-time — meaning cycles compress by a factor of λ ≈ 2.01 with each successive repetition (approximately an age-doubling). This is the signature of discrete scale invariance: the system looks the same not at every scale, but at a geometric sequence of scales.

The Hidden Coupling Constant C = k · ln(λ) ≈ 3.96: The two scale-invariances are not independent. Their relationship is governed by C, a single dimensionless constant that plays the role of a renormalisation-group invariant — the same mathematical structure that describes critical phenomena in statistical physics (phase transitions, percolation, fractal growth). C encodes how much the structural price floor rises with each completed cycle: exp(C) ≈ 52×. A market that finds its floor at $100 in one cycle finds it near $5,200 in the next.

Composite model fit: The chart fits five harmonic modes simultaneously via OLS on the Power Law residuals: the fundamental (m = 1) and harmonics m = 2, 3, 4, plus a hybrid mode at m = 3.5 identified by Perrenod as carrying significant power in the data. The composite model achieves R² ≈ 0.98 against raw log-prices — the remaining 2% variance is the true unpredictable component. The ±1σ prediction band is derived from the standard deviation of the fit residuals and widens in the outlook zone to reflect growing uncertainty beyond the data horizon.

Projection methodology: The composite model is evaluated forward to 2045 using the fitted parameters. The ~2028 peak annotation marks the next local maximum of the composite curve — not a fixed calendar target, but the point at which the current harmonic structure reaches its next crest given today's fit. The projection assumes parameter stability; any structural break (regulatory shock, technological disruption, change in adoption dynamics) would invalidate it.

Backtest reference: All quantitative claims on this page (median returns, win rates, activation counts, run lengths) derive from a single backtest run against daily close prices from 2010-07-18 to 2026-03-11 with the parameters documented above (CURRENT_WINDOW = 30 d, weights [0.5–3.0], HYSTERESIS_ON/OFF 55/45). If any of these parameters change, or if the backtest is re-run on a later dataset, all highlighted figures in this article must be reviewed and updated.

Limitations and Risks

  • Cycles are statistical patterns, not laws: Bitcoin can break from any identified regime at any time due to macro shocks, regulatory change, or structural market shifts.
  • Small effective sample sizes: The High and Extreme activity regimes, the Annual and Halving bands, and score crossover events all rest on fewer than 10 independent (non-overlapping) observations. Medians and win rates at these levels carry very wide uncertainty intervals — they describe what happened historically, not what will happen next.
  • Band count does not reliably rank returns: There is no consistent monotonic relationship between the number of simultaneously active bands and forward returns. The counter-intuitive ordering (0 bands = highest median 90-day return, +22.7%) means the Equalizer must be read as a structural description, not a return forecast. What does carry signal is the direction of change: a declining band count has historically been more reliably negative than a rising band count is positive (drops to 3–4 active bands: median 90-day −5% to −30%; fewer than 10 independent observations each).
  • Activation lags the move: Shorter bands (Short-term to Quarter) typically activate 40–61% above the prior 6-month trough; longer bands (Annual, Halving) activate 107–117% above it. Do not use activation events as precise entry or exit triggers.
  • Phase estimates carry uncertainty: "Days to trough" is a probabilistic model output. The uncertainty range (±30–50%) is shown directly in the interface. It is not a guaranteed date.
  • Look-ahead bias in backtested figures: The percentile ranking uses the full historical dataset as a reference. In a live setting this is correct by design; in the backtest it introduces a mild optimistic bias, meaning actual forward performance may differ from the quoted medians.
  • Scaling Law: in-sample fit and extrapolation risk: The Power Law and harmonic parameters are fit to the full available price history — there is no out-of-sample validation period. Five harmonic modes fitted to ~5,700 data points carry a meaningful risk of overfitting, particularly for the higher harmonics (m = 3, 4, 3.5). The outlook to 2045 is a model extrapolation over a horizon that exceeds Bitcoin's entire price history; uncertainty grows rapidly beyond the current data edge. The composite model has never been tested on a Bitcoin market that has undergone a fundamental structural change (e.g. sovereign adoption at scale, a critical protocol failure, or sustained regulatory suppression). Treat the projection as a structural hypothesis, not a price target.
  • Not financial advice: This analysis is informational only. Never rely on a single indicator as the sole basis for any investment decision.

Conclusion

Bitcoin moves in cycles — and the Bitcoin Cycle Monitor makes those cycles legible. Whether the market is quietly building structural energy or running hot in the historically significant High range, the monitor gives you a data-driven read in seconds. The Overall Market Activity score is the primary signal to watch: a reading in the High range has been the most consistent structural precursor to significant upward moves in Bitcoin's history. The Cycle Equalizer adds texture — which timescales are engaged, how long they typically persist, and how far along a cycle the market may be.

The Scaling Law chart adds a third dimension: structural position. The Wavelet tools measure rhythmic energy in the present; the Scaling Law measures where the present sits within a decade-long structure. Used together, they answer two distinct questions — how active the market is right now, and how far it has travelled from its long-run equilibrium. Neither question is more important than the other; both are necessary for a complete picture.

Markets change, and no signal works indefinitely. But understanding the rhythmic structure beneath price action — and knowing how unusual it is by historical standards — is a perspective worth returning to. Check back regularly: the monitor updates daily, and the structural picture can shift meaningfully within weeks.