Of the 20.022 million Bitcoin mined to date, only around 2 million are realistically tradable. The Long-Term Holder cohort grows from cycle to cycle, the freely available coin supply shrinks with each cycle. At the same time, the buyer base has structurally changed since the ETF launch in January 2024: less price-sensitive, more long-term oriented, larger in scale. What happens when this new demand profile meets a historically tight supply?
Executive Summary
As of late April 2026, 20.022 million of the maximum 21 million Bitcoin have been mined. Of these, Long-Term Holders (LTHs, defined as wallets that have not moved their coins for more than 155 days) hold roughly 16 million BTC — nearly 80 percent of total circulating supply. This includes an estimated 3 to 4 million permanently lost coins, 1.32 million BTC in US Spot Bitcoin ETFs, and around 818,000 BTC in Strategy's (formerly MicroStrategy) treasury alone. The trough of LTH holdings near the October 2025 peak stood at around 14.2 million coins — 2 million more than at the 2021 top, and over 5 million more than back in 2017.
What remains is a freely tradable float on the order of 4 to 6 million BTC. At a spot price near $78,000, that represents a liquid market value of roughly $300 to $450 billion — a fraction of what classical bond markets turn over on an average trading day.
The thesis of this piece: the market is approaching a structural point where classical demand-supply models no longer scale linearly. When institutional demand via regulated vehicles meets a supply that is over 70 percent in passive hands, non-linear price reactions are the likely outcome. This piece breaks down the data, examines historical parallels, lays out three scenarios, and names the counterarguments.
1. The Chart: What the LTH/STH Breakdown Shows
The starting point is a Glassnode classic, here in Checkonchain's rendering: Bitcoin Supply by Long-Term and Short-Term Holders, overlaid with spot price since 2010.
The central visual observation: the blue area (LTH supply) does not grow linearly but in steps. Each bull market redistributes coins from Long-Term Holders to Short-Term Holders — visible as compression of the blue area during price peaks in 2013, 2017, 2021, and 2024/25. In each subsequent bear market, LTHs accumulate again, each time on a higher absolute level than before.
The Cyclical LTH/STH Oscillation
The Bitcoin LTH cohort follows a clear pattern:
- Bear market + recovery: Buyers hold long → STH coins mature into LTH → LTH supply rises to a new high ("Peak HODL")
- Bull run to all-time-high: LTHs distribute into strength → LTH supply falls → STH supply rises
- After the bull peak: A new accumulation cycle begins on a higher floor than before
| Cycle Phase | Date (approx.) | LTH Supply (BTC) | STH Supply (BTC) | LTH Share |
|---|---|---|---|---|
| Peak HODL before Bull 2017 | Spring 2017 | ~12M | ~4M | ~75% |
| Trough at Bull Peak Dec 2017 | Dec 2017 | ~9M | ~7.4M | ~55% |
| Peak HODL before Bull 2021 | Oct 2020 | ~14.3M | ~4.2M | ~77% |
| Trough at Bull Peak May 2021 | May 2021 | ~12.2M | ~6.52M | ~65% |
| Peak HODL before Bull 2024/25 | Dec 2023 | ~16.4M | ~3.5M | ~83% |
| Trough at Bull Peak | Oct 2025 (est.) | ~14.1M | ~5.8M | ~71% |
| Current (Apr 2026) | Q2 2026 | ~16.1M | ~3.9M | ~81% |
Values rounded, based on Glassnode/Checkonchain data. The LTH highs before each bull run sit progressively higher than the previous cycle — a consistent structural observation across four cycles.
The key implication: the STH pool — the pool of younger, more frequently moved coins — has remained structurally near 3.5 to 6 million BTC for almost a decade, even though total supply has grown by over 40 percent during that time. Every newly mined Bitcoin statistically lands almost entirely in the LTH cohort. Put differently: issuance flows directly into passive hands.
2. The Math of the Float
If you define "float" as the supply realistically tradable in the short term — that is, coins available within weeks, not years, on exchanges or in custody structures — the breakdown as of April 2026 looks as follows:
| Category | Amount (BTC) | Share of Total Supply |
|---|---|---|
| Total mined | 20,022,000 | 100.0% |
| Estimated lost (Lost Coins) | −3,500,000 | −17.5% |
| Effectively existing | 16,522,000 | 82.5% |
| Long-Term Holders excl. lost coins (>155d) | −12,700,000 | −63.43% |
| Short-Term Holders (remainder) | 3,822,000 | 11.6% |
But even within the LTH category, the institutional share has grown sharply since 2024 — and that share is structurally non-recoverable for the free market:
| Sub-Category within LTH | Amount (BTC) | % of LTH Holdings |
|---|---|---|
| Estimated permanently lost coins | ~3,500,000 | 21.7% |
| US Spot Bitcoin ETFs (regulated custody) | 1,322,000 | 8.2% |
| Strategy Treasury (MSTR) | 818,000 | 5.1% |
| Other Public Treasury Companies (MARA, Riot, Tesla, Block, Metaplanet, BSTR, 21 etc.) | ~350,000 | 2.2% |
| Of which institutionally active (ETF + Treasuries) | ~2,490,000 | 15.5% |
| Remaining LTH (private wallets, OG cohort) | ~10,010,000 | 62.1% |
The institutional component within the LTH cohort has grown structurally from essentially zero to roughly 2.5 million BTC since the ETF launch in January 2024 — a rise to 15.5% of LTH holdings in just 28 months. These coins are unlikely to flow back into the STH pool, because ETF investors allocate strategically and treasury companies have built their equity story on "we never sell."
The real, short-term liquid float thus sits at roughly 2 million BTC — the part of the STH pool actively tradable on spot exchanges (the rest sits in private wallets of new holders that will soon cross the 155-day threshold and become LTH themselves).
That is the order of magnitude that the ETF´s and treasury companies accumulated in the last few years alone.
3. The Structural Shift on the Buyer Side
Historical Bitcoin demand was retail-driven. In the 2013, 2017, and 2021 bull markets, it was predominantly retail investors who entered late, sold in panic during drawdowns, and caused the typical 75-to-85-percent bear markets. Retail behavior is highly price-sensitive and reflexive: when price rises, demand rises; when it falls, demand collapses.
Since the ETF launch in January 2024, this dynamic hasn't disappeared — but it now competes with two new buyer categories whose behavior is fundamentally different.
3.1 ETF Investors
US Spot Bitcoin ETFs hold a cumulative 1.32 million BTC. The underlying buyer base consists primarily of wealth managers, RIAs (Registered Investment Advisors), and institutional allocators who hold Bitcoin as a 1-to-3-percent position in diversified portfolios. The decisive point: these allocations are typically calibrated by risk-parity or strategic asset allocation models, not by price level.
Concretely, this means: if a wealth advisor decides that 2 percent of client AUM should go into Bitcoin, buying happens — whether BTC is at $60,000 or $100,000. That is price-inelastic demand. It doesn't react to price; it reacts to the allocation decision. This is the same structural effect that fundamentally changed gold after the GLD launch in 2004.
3.2 Treasury Companies
The second category is even more radical. Treasury companies like Strategy (MSTR), Metaplanet, Bitcoin Standard Treasury Company (BSTR), and Twenty One (XXI) have established Bitcoin not as an investment position but as a balance-sheet strategy. Strategy under Michael Saylor is the textbook case: 818,000 BTC accumulated without selling since 2020, financed via convertibles and equity issuance. Average entry price is approximately $75,000 — at the current $78,000, the position is in the green, and Saylor keeps buying.
This buyer category has two properties that make it even more inelastic than ETF investors:
First: the logic of the "Bitcoin standard" as a treasury strategy is counter-cyclical. Treasury companies buy more aggressively during drawdowns because their marketing story rests on being "smart money." Strategy made its largest purchase since late 2024 on April 20, 2026 — exactly as BTC recovered from the January low at $60,000 and retail sentiment was still in bear mode.
Second: selling pressure is asymmetric. A treasury company can theoretically sell, but it would destroy the entire equity story. Strategy carries an equity premium on its NAV (mNAV at times above 2x, currently around 1.25x) that depends on the promise "we never sell." Selling would not only reduce Bitcoin holdings but collapse the share price — a classic bank-run scenario.
3.3 Comparison of Buyer Types
| Property | Retail | ETF Investor | Treasury Company |
|---|---|---|---|
| Price sensitivity | Very high | Low | Counter-cyclical (negative) |
| Average holding period | Weeks to months | Years | "Forever" |
| Selling triggers | Sentiment, FUD, panic | Allocation rebalancing | Balance-sheet crisis, liquidation |
| Share of trading volume | High (>60% pre-2024) | Medium | Minimal |
| Accumulation in drawdowns | Low | Medium | High |
4. What Happens in a Demand Shock on Illiquid Supply?
The theoretical framework comes from classical microeconomics: under perfectly inelastic supply, every additional unit of demand drives price upward at marginal costs that approach infinity. Under perfectly inelastic demand, the inverse holds.
Bitcoin today is approaching both extremes simultaneously: supply is algorithmically capped (450 BTC new per day at the current halving stage), and demand via ETFs/Treasuries is price-inelastic.
4.1 Three Scenarios
| Scenario | Trigger | Effect on Float | Price Implication |
|---|---|---|---|
| Mild (Base Case) | ETF inflows steady, treasury buys constant | Reduction 200-300k BTC/year | Structurally upward, low vol |
| Base | Allocation shift (Pension Funds, SWFs) | Reduction 500-800k BTC/year | Jump moves, squeeze phases |
| Extreme | Sovereign Buying + Banking crisis | Reduction >1M BTC/year | Reflexive liquidity crisis |
In the mild scenario currently underway, we see what played out in 2024 and 2025: phases of strong upward moves followed by consolidations as retail sellers (the STH pool) deliver into every rally. But the absolute level rises structurally because issuance is thin (165,000 BTC/year) and ETF/treasury demand exceeds that issuance several times over.
In the base case scenario, a larger allocation shift would occur — for example if US pension funds (TSP, CalPERS-equivalents) systematically allocate 0.5 to 1.0 percent into Bitcoin. With US pension AUM around $35 trillion, that alone equals demand of $175 to $350 billion spread over multiple years. With a float of $300-450 billion, that's mathematically not deliverable without price revaluation.
In the extreme scenario, states would become strategic buyers. El Salvador set the precedent. Bhutan, the US (Bitcoin Strategic Reserve announced under the Trump administration), and select Sovereign Wealth Funds could follow. If a single SWF the size of Norway's decided to hold 1 percent in Bitcoin, that would be $16 billion — roughly 4 to 5 percent of the current float. From a single buyer.
4.2 The Reflexivity of the Squeeze
A defining property of a real liquidity squeeze is reflexivity in the Soros sense: rising prices cause behavioral changes which in turn drive prices higher. In Bitcoin's context, the causal chain reads:
Step one: prices rise due to structural demand.
Step two: LTHs see their buy-and-hold thesis confirmed, hold further (instead of distributing at all-time highs).
Step three: the STH pool shrinks because no LTHs fall back into the STH category.
Step four: marginal purchases have larger price impact because less float is available.
Step five: treasury companies and sovereigns observe the price action, accelerate their allocations.
This sequence is not hypothetical. It already played out in 2024, when the ETF launch led to a doubling of BTC price within 12 months. The question is not whether the mechanism works, but whether it ignites again in a second and third wave.
5. Historical Parallels: Gold 2004, Silver 2021
History provides two useful comparisons for the ETF phenomenon.
Gold 2004: State Street's GLD ETF launched in November 2004, with gold at around $444/oz. By the September 2011 high, gold had reached $1,920/oz — a 4.3x move over 7 years. The driver was the same mechanism: wealth managers and institutional allocators got, for the first time, a regulated, easy-access vehicle. Before 2004, you had to buy physical gold or mining stocks — both with frictions that mirror today's Bitcoin custody hurdles. GLD removed those hurdles and opened a new buyer category.
The important difference: unlike Bitcoin, gold has no fixed supply cap. With rising prices, more gold can be mined, making issuance elastic. Bitcoin can't do that. This makes the mechanism theoretically stronger for Bitcoin, not weaker.
Silver 2021 (Reddit Squeeze): In January 2021, retail investors from WallStreetBets attempted a short squeeze in silver. The attempt failed because the silver market — unlike GameStop — is extremely deep and liquid (over $1 trillion in physical holdings). The comparison shows what doesn't happen when the float is too large. Bitcoin is the opposite: $300-450 billion in float, of which perhaps $50 to $80 billion is genuinely tradable on spot exchanges in the short term.
6. The Counterarguments
An honest analysis must also examine the other side. Three arguments deserve serious consideration.
First: LTH behavior can flip. The assumption that LTHs don't sell holds empirically — up to a price level that changes their life planning. If BTC runs to $250,000 or $500,000, LTHs will distribute. That has happened in every prior cycle. The difference is that the new LTHs (post-2020 entrants) have historically never sold in such a price range — we don't know whether they behave differently from the OG cohort.
Second: Treasury companies can be forced to sell. Strategy is refinanced via convertibles. If the share price falls below certain thresholds while convertibles mature, a cash crunch could emerge. With BSTR and 21, the situation is even more fragile because they start with leverage and lack the same equity premium buffer. A forced treasury sale during a bear-market phase could push 100,000 to 300,000 BTC into the free market — enough to noticeably loosen the float in the short term.
Third: Regulatory shocks. The ETF structure rests on SEC approval. A Trump administration is Bitcoin-friendly, but a future government could change that. Tax treatment, wash-trading rules, custody requirements — all three could slow or reverse the inflow dynamic. The argument is not that it will happen, but that the thesis rests on institutional frameworks that aren't politically guaranteed.
7. What This Means for Practice
The analysis leads to three frameworks for observing the coming quarters. They are not trading recommendations — Backtesting Arena is explicitly not an investment advisor. But they are testable hypotheses.
Framework 1 — Float tracking as a lead indicator: When Glassnode data shows STH supply falling below 4 million BTC (historical low since 2018, currently already at around 3.9 million), that's a quantifiable squeeze signal. This metric is integrable as a data feed in Backtesting Arena and can serve as a filter for long strategies.
Framework 2 — ETF/Treasury flow correlations: The hypothesis that cumulative ETF inflows + treasury accumulation has a 90-day lead on price action can be tested retrospectively. If the correlation holds (R² > 0.4 in out-of-sample data), it becomes a usable signal.
Framework 3 — Negative scenarios as a hedge: The counterarguments in Section 6 are not theoretical concerns but tail risks with measurable triggers. Strategy share price below $200 alongside falling BTC is a concrete warning signal for forced treasury sales. Such triggers can be hardcoded as risk-off conditions in strategies.
8. Conclusion
The data shows a clear structural trend: Bitcoin supply has been redistributed from short-term to long-term holders for over a decade. The free float is shrinking relative to total supply. At the same time, the demand side has fundamentally changed since 2024 — from price-sensitive retail to price-inelastic allocators and balance-sheet strategic buyers.
These two trends — shrinking float and inelastic demand — are not independent. They reinforce each other. When most of the issuance flows directly into passive hands, less material is available each year for the market clearing mechanism. The naive assumption that this simply translates into a higher price underestimates the non-linear properties of a real liquidity squeeze.
What the analysis does not do: it doesn't predict a price, doesn't predict timing, and is explicitly not to be read as investment advice. What it does: it identifies the structural vectors against which the coming quarters can be observed rationally — and which are testable.
That is exactly why Backtesting Arena exists.
Backtesting Arena makes Bitcoin strategies systematically testable.
Disclaimer: This article is not investment advice. It describes structural analyses based on publicly available on-chain and market data. Independent research and individual risk assessment are essential.
Data sources: Glassnode, Checkonchain, SIFMA Q1 2026 Fixed Income Quarterly.
FAQ EN:
Question: What does "Long-Term Holder" mean specifically?
Answer: Glassnode defines Long-Term Holders as wallets whose coins have not moved for more than 155 days. This threshold has proven statistically robust as an indicator for "conviction holders" — that is, holders who don't sell on normal price movements. Important: per Glassnode's definition, the LTH figure also includes the estimated 3 to 4 million permanently lost coins as well as holdings in spot ETFs and treasury companies, since all of these coins long ago crossed the 155-day threshold.
Question: Are the cited LTH figures historically verifiable?
Answer: Yes. Glassnode and Checkonchain provide publicly accessible data on the LTH/STH split going back to 2010. The Peak HODL and trough values used in this piece are derived directly from these datasets and reproducible. The cyclical oscillation — LTH highs before each bull run, lows at the bull peak — is a consistent pattern across four cycles.
Question: How do ETF buyers differ from earlier Bitcoin buyers?
Answer: Before 2024, Bitcoin buyers were predominantly retail and highly price-sensitive — they bought into rising prices and sold in drawdowns. ETF investors operate on strategic asset allocation: when a wealth advisor sets a 2% Bitcoin allocation, buying happens — regardless of price. This decoupling of price and demand is the central structural difference.