Bitcoin ETF Flows vs. The Halving Cycle: Which Actually Drives Price?
In 2024, spot Bitcoin ETFs launched and absorbed more than 12 times what miners produce daily. Multiple analysts declared the four-year halving cycle dead. We ran 10 years of data to test that claim. The result is more nuanced—and more useful.
Current Snapshot LIVE
The Central Question
In 2024, Bitcoin spot ETFs launched in the US and changed the demand landscape overnight.
By 2025, ETFs were absorbing more than $500 million per day in inflows—more than 12 times what Bitcoin miners produce daily post-halving (approximately 450 BTC per day at current issuance rates, or roughly $35M at prevailing prices).
The arithmetic sounds decisive: if institutions are buying 12x the new supply, surely they're now the dominant price force? Surely the 4-year halving cycle is a relic of retail-dominated markets?
Several prominent analysts made this exact argument heading into 2026. ARK Invest noted the ETF flow cycle had "replaced" the halving cycle. Grayscale predicted institutional adoption was "ending the traditional four-year cycle."
We ran the numbers. The reality is more nuanced—and more useful for investors.
What ETF Flows Actually Show
When spot Bitcoin ETFs launched in January 2024, they created a new demand structure that had never existed in Bitcoin’s history.
| Period | ETF Flow (BTC) | Mining Supply (BTC) | Absorption Ratio |
|---|---|---|---|
| May 2025 (peak) | +26,700 BTC/month | ~7,200 BTC/month | 3.7× |
| Peak inflow periods 2024 | $500M+/day | ~450 BTC/day | 12× |
| Nov 2025–Jan 2026 | −$6.18B net | ~450 BTC/day | Net distributor |
Academic research confirmed what traders observed: a study in the Ledger Journal found "periods of expanding Bitcoin ETF assets correspond to higher Bitcoin price levels, with cointegration confirmed at 10% significance level."
Short-term: flows follow price. Long-term: sustained flows drive price by absorbing available liquid supply. But this only tells half the story.
The Halving Baseline: What Our Factor Study Shows
Our factor hierarchy study covers 941 data points from January 2015 through December 2025. Using eta-squared from one-way ANOVA analysis, we measured exactly how much variance in Bitcoin’s 90-day returns each factor explains.
Factor Hierarchy (2015–2025)
The halving cycle phase is the dominant driver of Bitcoin’s 90-day returns. A factor that explains 28.2% of return variance is remarkably powerful in financial markets—the Fed rate regime, by comparison, explains only 9.4%.
Critically, this 28.2% figure was computed across the full dataset including the ETF era. The halving cycle didn’t lose explanatory power after ETF launch; if anything, the 2024–2025 bull phase followed the historical pattern closely.
The Plot Twist: ETFs Became Net Sellers
Here’s what the "ETFs ended the cycle" narrative missed: institutional demand can reverse.
From November 2025 through January 2026, Bitcoin spot ETFs experienced $6.18 billion in net outflows—including a record $4.57 billion in two months. For the first time, the vehicles designed to absorb Bitcoin supply became net distributors of it.
The mechanism was mechanical: with Bitcoin peaking near $126,198 in October 2025 then retracing sharply, institutional ETF holders found themselves underwater on average (average cost basis ~$90,200). This created selling pressure that reinforced the drawdown.
The 4-year cycle theory predicts distribution-phase underperformance. What happened in 2025–2026—a peak at roughly a +2σ power law event followed by a sharp drawdown—is entirely consistent with historical cycle patterns. ETFs amplified both the upswing and the downswing. They didn’t replace the cycle; they turbocharged it.
The Power Law as the Envelope
This is where BML’s model provides clarity that neither ETF bulls nor cycle skeptics fully capture.
The power law model describes Bitcoin’s long-term price corridor—the mathematical trend that has held for over a decade with R² = 0.9605. Historical sigma bands define where prices have reached at cycle peaks (+2σ to +3σ) and troughs (−1σ to −2σ).
Our hypothesis: ETF flows don’t change the corridor. They determine where within it price trades at any given moment.
- High, sustained ETF inflows → price approaches upper band
- ETF outflows combined with distribution phase → price tests lower band
- Neither moves price permanently outside the power law envelope
The October 2025 peak at $126,198 was approximately a +2σ event relative to the power law trend—entirely consistent with historical cycle tops, regardless of ETF involvement.
Where We Stand: April 2026
As of April 25, 2026, the data paints an interesting picture.
Current Cycle Position LIVE
We are 735 days post the April 2024 halving—deep in what our model classifies as the distribution phase (days 547–913). Historically, this phase has a 67.1% win rate and a +19.3% average 90-day return when prices are near fair value.
But here’s the notable data point: Bitcoin is currently trading at $77,640—approximately 44% below the power law fair value of $139,710. This is a rare configuration. The drawdown from the October 2025 peak ($126,198) has been amplified by institutional ETF outflows ($6.18B+ over November 2025–January 2026), and macro headwinds in early 2026.
Historically, trading this far below power law fair value during the distribution phase is the type of setup that has preceded meaningful recoveries. The factor model isn’t calling a bottom—but the valuation signal is as stretched as it has been since mid-cycle lows.
ETF flow direction is the confirmation variable. Watch whether institutional flows flip from net outflow to net inflow. When that happens during a period of deep undervaluation within the distribution phase, it historically signals strong forward returns. You’re looking for the same ETF amplifier that drove the upswing—this time off a compressed base.
The Synthesis: Flows Accelerate Halving Dynamics
After running the analysis, here’s the working model:
- Halving cycle sets the base probability. Our factor study confirms: 28.2% of 90-day return variance is explained by cycle phase alone. This didn’t change in 2024–2026.
- ETF flows are a demand amplifier, not a cycle replacement. Post-halving supply shocks were magnified by ETF absorption—the 95.8% historical bull phase win rate still held (and was arguably improved). But when institutional sentiment turned, the drawdown was similarly amplified.
- Power law defines the floor and ceiling. Neither ETF buying at 12x supply nor institutional selling has broken the long-term power law envelope.
Practical implication: The factor hierarchy (halving → valuation → macro) remains your primary framework. ETF flow direction is worth monitoring as a confirmation signal—strong inflows during bull phase = amplified upside. Outflows during distribution = amplified downside.
Conclusion
The "ETFs ended the halving cycle" narrative was intuitive but wrong.
The halving cycle still explains 28.2% of Bitcoin’s 90-day return variance—nearly three times the Fed’s influence. The power law corridor, with R² = 0.9605 over a decade, remains intact. The October 2025 cycle peak was precisely what the model predicted: a +2σ event in the late bull phase.
ETF flows changed the velocity and amplitude of Bitcoin’s moves, not the direction dictated by the cycle. They accelerated the upswing, then accelerated the correction. The fundamental framework—halving cycle, power law valuation, macro regime—held throughout.
As of April 2026, the data says: we’re in late distribution phase, priced 44% below fair value, with ETF flows as the key confirmation signal for what comes next.
Track the Signals in Real-Time
Our free dashboard shows current cycle phase, power law deviation, and ETF flow direction—updated daily. No narrative. Just data.
View Live DashboardMethodology
Factor importance data covers January 2015–December 2025 (941 observations). Factor importance measured using one-way ANOVA with eta-squared as the effect size metric. 90-day forward returns used as the dependent variable. Halving phases defined by days elapsed since last halving event, with boundaries at days 0, 547, and 913. Power law valuation fair value calculated using the diminishing sine power law model (R² = 0.9605). Live price data from CoinGecko/Coinbase public API (last updated April 25, 2026). ETF flow data referenced from publicly reported figures (BitMEX Research, Farside Investors).
Disclaimer: This is statistical analysis of historical patterns, not financial advice. Past patterns do not guarantee future results. Always do your own research.
Related: Factor Hierarchy: What Actually Moves Bitcoin Price? • Bitcoin Halving Cycle Guide • Whale & Institutional Flows