Log-periodic cycles and the Perrenod-Santostasi wave model

The Perrenod-Santostasi log-periodic wave model extends the Power Law trajectory with a discrete-scale-invariance overlay — sinusoidal oscillations in log-time that decompose Bitcoin's cycles into modes around the continuous-scale-invariance Power Law trend. The empirically-fit ratio is $\lambda \approx 2.01$, producing fundamental cycle peaks when Bitcoin's age doubles — roughly 2010, 2011, 2013, 2017, and 2024-2025. The framework's distinguishing claim is that the 2021 peak was a harmonic at $\sqrt{\lambda }\approx 1.4\times$ spacing, not a fundamental — explaining its anomalous structure. Log-periodic modes capture about 74% of the variance residual to the Power Law trend; the coupling constant $C=k\cdot \ln (\lambda )\approx 3.97$ ties Power Law and discrete-scale-invariance together. The next predicted fundamental peak is mid-2028. Published through the Scientific Bitcoin Institute; draws on Didier Sornette's broader LPPL literature on financial-market critical phenomena.

---

Why this note matters

The log-periodic framework is the most analytically substantive cyclical alternative to the four-year halving cycle. Three reasons it is load-bearing.

First, it is the rigorous theoretical alternative to halving-as-master-cycle. The four-year framework is largely descriptive; the log-periodic framework provides a mathematical structure — discrete scale invariance coupled to the Power Law — that predicts cycle structure from first principles. Four-year observations become one cycle structure among many, not the master pattern.

Second, it explains the 2021 anomaly cleanly. The four-year framework struggles with the cycle’s double-peak structure and the deep 2022 drawdown. The log-periodic framework treats 2021 as a predicted harmonic, not a fundamental — with the next fundamental in mid-2028. This resolves the four-year framework’s biggest empirical difficulty.

Third, it integrates with the Power Law derivation. The coupling constant $C=k\cdot \ln (\lambda )$ ties the Power Law exponent to the discrete-scale-invariance ratio, producing a unified trajectory-plus-cycle model rather than two separate frameworks.

For the foundational Power Law trend, see The Power Law model; for the framework this note critiques, see Four-year halving cycles.

---

The mathematical framework

Continuous vs. discrete scale invariance

The Power Law model (The Power Law model) is a continuous scale invariance framework — $P(t)\propto t^n$ has the property that for any scale factor $s$, $P(s\cdot t)=s^n\cdot P(t)$. The structure looks the same at every scale.

The log-periodic framework adds a discrete scale invariance overlay — for a specific scale factor $\lambda$, $P(\lambda \cdot t)={\lambda }^n\cdot P(t)$ but for intermediate scales the structure shows distinct modes. The “discreteness” of the invariance produces cyclical structure: features at scale $t$ are correlated with features at scale $\lambda t$, ${\lambda }^{2}t$, etc., but not at intermediate scales.

Empirically for Bitcoin: the discrete-scale-invariance ratio is $\lambda \approx 2.01$ — meaning Bitcoin’s structure repeats when its age doubles. Major peaks occur at ages following the doubling pattern: 1, 2, 4, 8, 16 years.

The mathematical signature: oscillations in $\ln (t)$ rather than $t$. The cycles aren’t periodic in calendar time; they are periodic in log-time.

The LPPL functional form

The full log-periodic-power-law (LPPL) functional form, expressing Bitcoin’s price as the Power Law trend plus log-periodic oscillations:

$P(t)=A\cdot t^n\cdot \left[1+{\sum }_{m=1}^M{B}_m\cdot \cos \left(m\cdot \omega \cdot \ln (t)+{\varphi }_m\right)\right]$

Where:

• $A\cdot t^n$ is the Power Law trend (continuous scale invariance)
• $m$ indexes modes — $m=1$ is the fundamental, $m=2,3,...$ are harmonics
• $\omega =2\pi /\ln (\lambda )$ is the fundamental angular frequency in $\ln (t)$
• $B_m$ is the amplitude of mode $m$
• ${\varphi }_m$ is the phase of mode $m$

The empirical fit produces specific values:

• Power Law exponent: $n=k\approx 5.68$
• Discrete-scale-invariance ratio: $\lambda \approx 2.01$
• Fundamental frequency: $\omega =2\pi /\ln (2.01)\approx 9.02$ radians per unit of log-time
• Mode amplitudes and phases: fit empirically from historical Bitcoin data

Modal decomposition (per Perrenod’s analyses) finds that about 74% of the residual variance around the Power Law trend is captured by the log-periodic modes — a substantial fraction supporting the framework’s empirical content.

The coupling constant

The framework derives a coupling constant linking the Power Law trend and the discrete-scale-invariance structure:

$C=k\cdot \ln (\lambda )\approx 5.68\cdot \ln (2.01)\approx 5.68\cdot 0.698\approx 3.97$

This coupling constant is a physical-system parameter: it relates the power-law growth rate (how fast the trend rises) to the discrete-scale-invariance spacing (how cycles map in log-time). Perrenod’s interpretation: the coupling constant is a property of Bitcoin’s specific system, analogous to how coupling constants in physics relate different fundamental forces.

The numerical value $C\approx 3.97$ has no obvious physical interpretation in the way other coupling constants do, but its consistency across Bitcoin’s history is itself a substantive empirical claim. If the framework continues to hold, the coupling constant should remain stable.

Background: Sornette’s log-periodic-power-law literature

The LPPL framework has a long history in financial-market analysis. Didier Sornette (Swiss-French physicist, ETH Zurich) and collaborators developed the framework extensively in the 1990s-2000s to model financial bubbles and crashes:

• Sornette and Johansen, “Critical Crashes” (1999) — foundational LPPL paper for financial markets
• Sornette, Why Stock Markets Crash (Princeton University Press, 2003) — book-length treatment
• Various subsequent papers applying LPPL to specific markets and crashes (1987, 2000, 2008, COVID, etc.)

Sornette’s LPPL framework treats financial bubbles as critical phenomena — phase-transition dynamics analogous to physical-system phase transitions. The framework predicts specific log-periodic signatures preceding bubble peaks.

Santostasi’s Bitcoin application (circa 2019) drew explicitly on Sornette’s earlier framework but with substantial Bitcoin-specific adaptations. Perrenod’s recent extension (2024-2026) has further developed the framework with the coupling-constant formulation and the mechanistic-derivation paper.

The relationship to Sornette’s framework matters because:

• It situates the Bitcoin application in a broader scientific tradition rather than treating it as Bitcoin-specific curve-fitting
• It provides external validation — LPPL signatures have been observed in many financial markets, not just Bitcoin
• It connects Bitcoin’s cycles to the broader critical-phenomena literature in physics and complex systems

---

Bitcoin’s specific log-periodic structure

Applying the framework to Bitcoin’s empirical history:

Fundamental cycle peaks (m = 1)

Per the doubling pattern ($\lambda \approx 2.01$), fundamental cycle peaks are predicted at Bitcoin ages following geometric progression. Each subsequent peak is at approximately twice the prior peak’s age:

Age	Calendar	Status
~1 year	2010	Initial price formation; rough fundamental
~2 years	2011	$32 peak — first clear cycle peak
~4 years	2013	$1,200 peak — second cycle peak
~8 years	2017	$19,500 peak — third cycle peak
~16 years	2025	~$124K peak (Aug 2025) — the predicted age-16 fundamental
~19.5 years (with phase consideration)	mid-2028	Next predicted fundamental peak
~32 years	2041	Subsequent predicted fundamental peak

The pattern holds remarkably well through the first four fundamental peaks (with appropriate allowance for the imprecision of “Bitcoin’s age” given early-history price-formation dynamics). The mid-2028 prediction is the live empirical test.

Harmonic peaks (m > 1)

Harmonics occur at intermediate scales between fundamental peaks. The empirical analysis identifies:

• $\sqrt{\lambda }\approx 1.414$ spacing: harmonics at $1.414\times$ each fundamental’s age
• For the age-8 fundamental (2017): a harmonic at age $8\cdot 1.414=11.3$ years (early 2020)
• For the age-16 fundamental (2024-2025): a harmonic at age $16\cdot 1.414=22.6$ years (mid-2031)
• For the age-12 region (corresponding to 2021): the framework identifies a strong harmonic from the m=2 mode

The 2021 peak at age ~12 years is the framework’s most distinctive prediction: this is not a fundamental cycle peak but a harmonic — explaining its anomalous post-peak drawdown structure.

Variance decomposition

Perrenod’s analyses report:

• Power Law trend explains: ~70-80% of total variance in log price
• Log-periodic modes explain: ~74% of residual variance (i.e., variance around the Power Law trend)
• Remaining unexplained variance: ~5-10% of total variance — attributed to noise, regime-change events, and modeling limitations

The combined fit is substantially better than the Power Law alone, supporting the framework’s empirical content.

---

Predictions

The framework generates specific, testable predictions:

Near-term: the 2025-2028 cycle

The four-year framework predicts: cycle peak in late 2025 or 2026; trough in 2026-2027; next cycle starts ~2027-2028.

The log-periodic framework predicts: no fundamental peak in late 2025 or 2026 — current dynamics are part of a long buildup toward the mid-2028 fundamental peak. Minor harmonics may produce localized cycle structure but not a full cycle peak.

Resolution: the 2025-2027 window is the empirical test. If a clear cycle peak appears in late 2025 or 2026 with substantial subsequent drawdown, the four-year framework is reinforced. If the trajectory continues approximately along the Power Law trend with no clear peak through 2027, the log-periodic framework is reinforced.

Medium-term: mid-2028 and beyond

The framework predicts a fundamental cycle peak in approximately mid-2028 (Bitcoin age ~19.5 years, factoring in phase considerations). Expected dynamics:

• Substantial appreciation through the buildup period (2026-2028)
• Cycle peak in mid-2028
• Substantial drawdown post-peak (consistent with prior fundamental cycles)
• Multi-year consolidation before the next fundamental cycle at age ~32 years (2041)

Coupling constant stability

The framework predicts the coupling constant $C\approx 3.97$ remains stable across cycles. Empirical violation (substantial drift in the coupling constant fitted across cycles) would be evidence against the framework’s structural content.

Higher-mode harmonics

The framework predicts specific harmonic structure: $\sqrt{\lambda }\approx 1.4\times$ spacing harmonics between fundamentals. Specific harmonic predictions can be derived for the 2026-2028 buildup period; deviations would inform framework calibration.

---

Empirical assessment

The framework’s empirical track record:

Supporting evidence:

• Fits the first four fundamental peaks (2010, 2011, 2013, 2017) with appropriate calibration
• Explains the 2021 anomaly — the harmonic interpretation accounts for the cycle’s double-peak and deep drawdown
• 74% residual variance decomposition is a substantial fraction
• Mechanistic connection to Power Law — the coupling constant provides theoretical structure
• Connection to Sornette’s broader LPPL framework — independent validation across many markets

Open empirical questions:

• 2025-2028 cycle behavior: framework predicts no near-term fundamental peak; four-year framework predicts a peak. The next 2-3 years resolve this
• Coupling-constant stability: across cycles, does $C\approx 3.97$ hold?
• Mode-amplitude attenuation: do harmonic amplitudes shrink consistent with the framework’s predictions, or do they vary in framework-inconsistent ways?
• Calibration of $\lambda \approx 2.01$: is this stable or does it drift? Some analyses suggest slight evolution

Where the framework is incomplete:

• Regime-change events: substantial macro or regulatory regime changes may produce dynamics the framework doesn’t predict
• Late-stage saturation: as Bitcoin approaches significant fractions of global wealth, the framework’s natural dynamics may break down
• Specific mode-amplitude evolution: the framework predicts the structure but the specific amplitudes are empirically fit rather than theoretically derived

---

Comparison with the four-year cycle framework

Dimension	Four-year cycle	Log-periodic
Cycle anchor	Halving events (every 4 years)	Discrete scale invariance ($\lambda \approx 2.01$, age-doubling)
Cycle structure	Single-peak per cycle (with 2021 anomaly)	Fundamentals + harmonics; multi-modal
2021 peak treatment	Anomalous double-peak; needs ad-hoc explanation	Harmonic at age ~12, explicitly predicted
Causal mechanism	Supply-shock from halvings (or adoption-narrative)	Critical-phenomena dynamics; discrete-scale-invariance in adoption
Power Law integration	Cycles oscillate around Power Law trend	Cycles ARE the log-periodic decomposition of Power Law residual
Predictive specificity	Cycle peak ~12-18 months post-halving	Specific predicted fundamental peaks at age-doubling intervals
Next cycle peak prediction	Late 2025 or 2026	Mid-2028
Mathematical rigor	Descriptive	Rigorous (LPPL formalism)
Empirical fit (residual variance)	Vague	~74% of residual decomposed
Background tradition	Bitcoin-specific empirical observation	Sornette’s broader LPPL literature; financial-market critical phenomena
Status	Standard framework; widely cited	Substantive alternative; not yet mainstream

The frameworks are fundamentally different ways of organizing Bitcoin’s cyclical structure. The four-year framework treats halvings as causal anchors and cycles as semi-discrete events. The log-periodic framework treats cycles as continuous decomposition of price residuals into modes around a smooth trend.

The empirical test in 2025-2028 will inform which is the more useful framework going forward.

---

Implications for allocation and trajectory

The framework’s allocation implications differ from the four-year-cycle framework’s:

If the log-periodic framework holds:

• The 2025-2027 period is buildup, not cycle: no substantial cycle-top distribution opportunity; long-hold dominates
• Mid-2028 is the next fundamental cycle peak: appropriate point for potential cycle-aware partial distribution
• The 2028-2032 period is the cycle drawdown: appropriate accumulation window for long-horizon allocation
• The 2041 fundamental is the next major peak: extremely long-horizon allocation thinking

If the four-year framework holds instead:

• 2025-2026 is cycle peak: cycle-aware partial distribution opportunity now/soon
• 2026-2027 is cycle drawdown: accumulation window
• 2028-2029 is next cycle buildup: long-hold through next cycle

The frameworks have different timing implications but similar long-horizon trajectory expectations. Investors who default to long-hold strategies are largely indifferent to which framework holds; investors attempting cycle-timing should explicitly consider both frameworks rather than relying on either alone.

For allocation framework integration, see Portfolio approaches to Bitcoin.

---

Counter-arguments and tensions

Discrete scale invariance may be over-fitting

The argument: The LPPL framework has many parameters ($\lambda$, multiple mode amplitudes and phases). Fitting Bitcoin’s price history with that many parameters will inevitably produce strong apparent fits. The 74% residual variance decomposition may be artifact of flexibility rather than substantive content.

Response: Substantive concern. The defenses:

1. The framework was specified before the test data (2024-2026 data is now out-of-sample for fits made earlier); continued fit is meaningful
2. Sornette’s broader LPPL literature provides independent validation that LPPL structure appears in financial markets generally
3. The coupling-constant derivation is a specific structural claim, not just curve-fitting
4. The 2025-2028 prediction is testable — if the framework continues to hold through the next cycle, that’s substantive evidence against pure over-fitting

The honest reading: over-fitting is a real concern; the framework’s predictive content over the next 2-3 years will inform whether the fit reflects structure or flexibility.

The 2021 peak interpretation is suspicious

The argument: The framework’s claim that 2021 was a “harmonic” rather than a fundamental looks suspicious — it conveniently explains away the cycle most damaging to standard frameworks (four-year, S2F). The choice of harmonic interpretation may be ex-post rationalization rather than ex-ante prediction.

Response: Partially fair. The framework’s harmonic interpretation of 2021 was developed alongside (and partly motivated by) the 2021 cycle’s anomalous structure. However:

1. The harmonic at $\sqrt{\lambda }$ spacing is a structural prediction of the framework, not a free parameter — given the framework’s $\lambda$ and mode structure, harmonics are predicted to occur
2. The 2021 timing fits the harmonic prediction, not just any-explanation prediction
3. Other cycles (2020 minor peak, etc.) also have harmonic interpretations that pre-existed the 2021 event

The honest reading: the harmonic interpretation has substantive content, but the framework’s robustness depends on continued empirical fit, not on retrospective explanation of past cycles.

Sornette’s broader LPPL framework has mixed empirical record

The argument: Sornette’s LPPL framework, applied to financial markets generally, has had a mixed predictive record. Some bubble predictions have been correct (some COVID-era predictions, some specific commodity predictions); others have not. The framework is not unambiguously validated even in its broader context.

Response: Fair. Sornette’s framework is taken seriously in critical-phenomena and financial-physics communities but is not consensus in mainstream finance. Bitcoin’s specific case may benefit from the framework being right about Bitcoin even if it’s mixed about broader markets. Or Bitcoin’s apparent fit may be coincidental. The honest reading: the framework deserves engagement based on its specific Bitcoin fit; the broader LPPL track record provides background context rather than independent validation.

The four-year cycle has strong intuitive grounding

The argument: The four-year cycle is intuitively grounded in halvings, mining economics, and observable cycle dynamics. The log-periodic framework requires accepting discrete-scale-invariance in $\ln (t)$ — a much more abstract and less intuitive mechanism. Occam’s razor favors the simpler framework when empirical evidence is ambiguous.

Response: Real consideration. The defense is that:

1. Intuitive grounding isn’t necessarily right grounding — many empirically correct frameworks are counterintuitive
2. The four-year framework has its own intuitiveness costs — the 2021 anomaly requires ad-hoc explanations
3. Empirical performance should weigh more than intuitive grounding — the log-periodic framework’s specific empirical fit is the relevant evidence
4. The frameworks are not mutually exclusive — halvings could trigger log-periodic modes; the integration is at least conceivable

The honest reading: intuitive parsimony favors the four-year framework; empirical content has been more favorable to the log-periodic framework recently. The choice depends on weighting.

Predictions are not testable on a short enough timescale

The argument: The framework’s specific predictions (mid-2028 fundamental peak) are several years out. The framework is essentially un-falsifiable on near-term timescales. By the time the prediction is testable, the framework may have been updated to fit whatever happened.

Response: Fair. The framework’s near-term content is limited; testing it requires multi-year time horizons. The defense is that science-of-financial-markets often requires such horizons. The framework should be cited with appropriate awareness that its testability is on multi-year timescales, and intermediate cycle dynamics may not provide clear signals.

Within-Bitcoin: the framework’s complexity may obscure rather than illuminate

The argument: The log-periodic framework requires substantial mathematical and physics background to engage rigorously. For most Bitcoin investors and analysts, the framework is too complex to operationalize. Its complexity may produce confusion rather than insight — and may not actually improve practical allocation decisions over simpler frameworks.

Response: Fair as a practical critique. The framework’s contribution may be substantively in long-horizon trajectory thinking (where Power Law alone suffices) rather than in cycle-timing (where simpler frameworks may be more accessible). The honest reading: log-periodic offers analytical sophistication that may be valuable for serious modelers and may be unnecessary for general allocation decisions.

---

Open questions for further development

• How will the 2025-2028 cycle resolve? This is the framework’s most important live test.
• Is the coupling constant $C\approx 3.97$ stable across cycles, or does it evolve? Stability would strengthen the framework’s claim to structural content.
• Are there higher-order modes (m = 3, 4, …) that fit Bitcoin’s price residuals? Higher-mode fits would either strengthen the framework (consistent structure) or weaken it (over-fitting).
• How does the framework engage Bitcoin’s institutional-adoption regime change? Institutional dynamics may produce different inflow patterns than retail-driven dynamics, potentially affecting the log-periodic structure.
• What is the relationship to Sornette’s broader LPPL framework? Does Bitcoin’s specific case provide insight into the general LPPL framework’s validity?
• How does the framework integrate with on-chain cycle indicators? On-chain data provides higher-resolution cycle information that could either support or contest the log-periodic decomposition.
• What is the appropriate timescale for the framework’s eventual breakdown? At very high adoption, all cyclical frameworks must eventually break down; when does the log-periodic framework transition into late-stage saturation?
• Does the framework have implications for Layer 2 / Lightning value flows? If Bitcoin’s value flows shift to Layer 2, base-layer cyclical structure may evolve.

---

Canonical sources for this note

Foundational LPPL literature

• Didier Sornette and Anders Johansen, “Critical Crashes” (Risk Magazine, 1999) — foundational paper applying LPPL to financial markets
• Didier Sornette, Why Stock Markets Crash: Critical Events in Complex Financial Systems (Princeton University Press, 2003) — book-length treatment
• Didier Sornette, various academic papers through 2000s-2010s elaborating LPPL framework
• Various subsequent papers applying LPPL to specific market crashes (1987, 2000, 2008, etc.)

Santostasi’s foundational Bitcoin LPPL work

• Giovanni Santostasi, “The Bitcoin Power Law Theory” (Medium) — includes log-periodic extension circa 2019
• Various Santostasi Substack posts and Twitter threads developing the framework
• Santostasi presentations at Bitcoin and physics-adjacent conferences

Perrenod’s recent LPPL extensions

• Stephen Perrenod, “Bitcoin’s Scaling Law: Power Laws, Log Periodicity, and a Hidden Coupling” (Substack) — the coupling-constant derivation
• Stephen Perrenod, “Disproving 4-Year Cycle Dominance in Minutes” (Substack) — the direct critique of halving-as-master-cycle
• Stephen Perrenod, “A Proposed Grand Unified Theoretical Framework for Bitcoin” (Substack) — the broader framework integration
• Stephen Perrenod, various other Substack writings developing the framework

Scientific Bitcoin Institute integration

• Santostasi and Perrenod, “A Mechanistic Derivation of the Bitcoin Price Power Law: Network Adoption Dynamics and Generalised Metcalfe Scaling” — the SBI paper integrating LPPL with the Power Law derivation
• Scientific Bitcoin Institute — institutional anchor

Background mathematical and physics literature

• Various critical-phenomena and discrete-scale-invariance literature from physics
• Various complex-adaptive-systems literature on power-law dynamics
• Various financial-physics academic literature on log-periodic structure

---

Related notes

• The Power Law model — the trend framework that log-periodic structure decomposes
• Four-year halving cycles — the standard framework this note explicitly critiques
• Adoption curves — adoption-side framework that may produce log-periodic-style cycles through cohort dynamics
• Metcalfe’s Law applied to Bitcoin — network-value framework underlying Power Law
• Stock-to-flow model — alternative supply-side framework; price-model engaged critically
• Diminishing returns thesis — cycle-attenuation framework consistent with log-periodic mode-amplitude attenuation
• Logarithmic regression and rainbow charts — precursor framework; less sophisticated cyclical handling
• Lindy effect and Bitcoin — survival framework adjacent to long-term trajectory
• The halving - Mechanism — supply-schedule mechanism; relationship to log-periodic structure is debated
• Bitcoin and global liquidity — macro-cyclical alternative framework
• Bitcoin and the ISM PMI cycle — alternative cyclical framework
• Portfolio approaches to Bitcoin — practical allocation implications of cyclical-positioning frameworks
• Long-term price models and cycles — sub-MOC for the price-models area
• Giovanni Santostasi — log-periodic framework originator; Power Law model originator
• Stephen Perrenod — log-periodic framework extension; coupling-constant derivation; SBI paper co-author
• Plan B — S2F framework; engaged critically
• Dylan LeClair — market-cycle analyst engaging cycle frameworks
• James Check — on-chain analyst engaging cycle positioning
• Ryan - On-Chain Mind — on-chain analyst engaging cycle positioning
• Scientific Bitcoin Institute — institutional anchor