Every Bitcoin cycle in the past decade has rhymed. Halving, slow accumulation, breakout, parabolic phase, blow-off top, deep bear. The pattern was reliable enough that traders built entire frameworks (Plan-B's stock-to-flow, the four-year cycle thesis, on-chain valuation models) on top of it. The pattern is also why most analyses of this cycle have been wrong: the 2024-2026 cycle has rhymed in some ways and broken in others, and the parts that broke are the ones that matter for what comes next.
This article walks through what is structurally different about the current cycle, what is the same, and what the on-chain and flow data is currently signalling about the back-half. It is not a price-target piece — anyone who tells you they know what BTC will print at year-end is selling something. It is a framework for sizing positions through a cycle whose shape is still being established.
The 2024 spot-ETF arrival changed the marginal buyer
January 2024's approval of US spot Bitcoin ETFs brought a new buyer cohort into the BTC market: registered investment advisors, family offices, and corporate treasuries that could not previously hold Bitcoin in a compliant wrapper. The flow signature of this cohort is fundamentally different from prior-cycle buyers.
First, ETF buyers are price-elastic in a different way. Retail crypto-native buyers in 2021 chased momentum — flows surged on the way up, dried up on the way down, and turned outright negative in drawdowns. ETF buyers have shown the opposite tendency: net inflows have been sustained during 10-15% drawdowns and have actually accelerated on some of them. The interpretation is that ETF buyers are executing pre-decided allocation plans rather than reacting to price action.
Second, ETF buyers are sticky. Once a wealth manager has onboarded BTC into a model portfolio at a 1-3% allocation, the position rebalances mechanically rather than getting actively traded. This sticky cohort has absorbed roughly 5-7% of total Bitcoin supply by mid-2026, depending on whose accounting you trust. That is a significant portion of the float that is unlikely to come back to market on a normal price swing.
Third, ETF flows print on a daily lag and are visible to everyone. This has created a new short-term feedback loop: BTC dips, ETFs print inflows the next day, traders take that as confirmation and add length, BTC bounces. The mechanic is real and has been a consistent feature of the 2024-2026 tape. It also creates the risk of a sharp unwind if the inflows ever pause for an extended stretch.
The halving still mattered, just less directly
Bitcoin's April 2024 halving cut the per-block subsidy from 6.25 to 3.125 BTC. The historical halving narrative was that this supply cut would mechanically pressure price higher 6-12 months later, as the daily supply absorbed by exchanges no longer matched buyer demand at prevailing prices.
That mechanic still operated this cycle but mattered less than in 2016 and 2020. Why: the absolute size of the daily subsidy cut (roughly 900 BTC/day pre-halving to 450 BTC/day post-halving) was small relative to other flows — particularly ETF inflows, which were running 5,000-15,000 BTC/day at peaks. The halving was a marginal supply effect; the ETF was a step-change in demand. The combination produced a stronger move than either alone, but the halving narrative on its own would have under-predicted the cycle's first leg.
What the halving did do unambiguously: it forced the marginal miner out of the business at the post-halving hash price. Mining sector profitability compressed sharply through mid-2024 before recovering as price rose. The pattern is identical to 2016 and 2020 and is one of the more reliable Bitcoin cycle features.
On-chain supply: long-term holders held, then sold, then held again
Glassnode and Coin Metrics both publish a metric called LTH supply — the share of BTC held in addresses that haven't moved their coins for 155+ days. Long-term-holder behaviour has been one of the cleanest leading indicators in past cycles: LTH supply tends to peak before price tops (long-term holders distributing into strength) and trough before price bottoms (long-term holders accumulating).
This cycle has shown the expected pattern but with timing shifts. LTH supply peaked in early 2024, declined steadily through the first leg of the rally as long-term holders distributed into ETF demand, troughed in mid-2025, and has since been recovering. The recovery in LTH supply through late 2025 and into 2026 has been interpreted as a sign that the cycle is in its mid-phase rather than the end-stage distribution phase that usually precedes a major top.
This is suggestive but not definitive. Past cycles' LTH peaks were tighter signals because there were fewer offsetting flows. With ETF and corporate treasury demand running concurrent with LTH distribution, the cohort signal needs to be read in context of the absorption side.
Mining economics: hash price compressed, then expanded
Hash price (revenue per terahash per second per day) is the right single number to track miner economics. It collapsed roughly 60% in the weeks after the halving as subsidy revenue halved while network hash rate kept growing. The recovery since has been driven by price appreciation outweighing further hash-rate growth and, intermittently, by transaction-fee spikes during congested mempools.
Why this matters for traders: the publicly-listed miners (Marathon, Riot, CleanSpark, etc.) function as a leveraged play on BTC price. When BTC rallies, miner equity rallies more, and when BTC pulls back, miners get hit harder. Their willingness to sell freshly-mined BTC into spot is also a function of their balance sheet — when prices are well above all-in cost, miners can hold; when prices compress, they have to sell. This is a small flow on net but consequential for short-term order-flow.
Volatility regime: lower, but with the same tail risk
Realised volatility on BTC has compressed through this cycle. The peak 90-day realised vol in 2024 was meaningfully lower than the equivalent peaks in 2017 or 2021. The interpretation is that ETF demand provides a buying cushion that absorbs intraday moves, and the deeper options market provides hedging supply that dampens implied vol.
What hasn't changed: the tail-risk distribution. BTC has had multiple single-day drops of 8-15% in the past 18 months, often on liquidation cascades that started in derivatives markets. The lower average vol does not extend to the tails. Position-sizing models calibrated to the lower mid-cycle vol will under-size against the tail risk.
Correlation regime: less tech, more macro
BTC's rolling correlation with NASDAQ tech was extremely high through the 2022 bear (BTC traded as a tech-stock proxy) and remained elevated through early 2023. It loosened in 2024 with the spot ETF and has been more episodic since. In risk-off episodes BTC still sells off with tech, but in mid-week macro-driven moves (FOMC, CPI, Treasury auctions) BTC sometimes moves independently of equity indices.
The cleaner correlation has been with the dollar index (DXY). BTC has shown a moderate negative correlation to DXY through the cycle: dollar weakness has been a tailwind for BTC; dollar strength has been a headwind. The relationship is noisy but consistent enough that DXY direction is one of the cleaner cross-asset tells on a multi-week window.
Where we are in the cycle
On every traditional cycle metric — LTH supply recovery, modest speculator positioning, miner economics improving rather than maxed-out, ETF flows still net positive — the data is consistent with a mid-cycle position rather than an end-cycle distribution phase. The classic blow-off-top pattern (parabolic price action, mass retail FOMO measurable in Google trends and exchange new-account signups, alts dramatically outperforming BTC) has not happened yet.
That does not mean it cannot still happen. Several past cycles had their largest move concentrated in the last 90-120 days of the cycle. If this cycle follows the classical four-year rhythm, the late-cycle phase would be expected somewhere in late 2025 to mid-2026 — which is now. The data is currently more consistent with "middle" than "late", but the gap between those two interpretations is small and the data revises week by week.
Practical sizing through the back-half
- Treat the structural ETF bid as a real but bounded buffer. It absorbs 5-15% drawdowns; it does not absorb 30-50% drawdowns.
- Watch the LTH-supply trend monthly. A sustained reversal to declining LTH supply would be the strongest single indicator that the cycle's distribution phase has begun.
- Don't size to the average volatility — size to the tail. 8-15% single-day moves remain part of the distribution despite lower realised vol.
- Cross-check on DXY for cleaner direction-of-the-day signal than the equity correlation provides.
- If altcoin season starts (BTC dominance falling sharply, alt aggregate outperforming BTC for multiple weeks), that has historically been a late-cycle marker — not necessarily the top, but consistent with the back half of the run.