The straight-line trap
More demand, higher price – until it isn't
We carry a simple model in our heads: more buyers means a higher price, and the line climbs smoothly. For most goods that intuition is roughly right, because supply quietly expands to meet demand. Coffee gets dear, more coffee gets planted. The line bends, but it bends gently.
Bitcoin breaks the model in two places. First, supply cannot expand – it is fixed at twenty-one million, with new coins arriving on a schedule no amount of demand can hurry. Second, and less obvious, the existing coins are arrayed on a ladder of reservation prices, and only the bottom rung is on offer cheaply. A thin sliver of coins is for sale near today's number; above it sit holders who would part with a coin at twice the price, fewer who would at five times, fewer still who treat theirs as a multi-generational inheritance and name no number they expect to see. So the price you watch tick up and down is not the price of all the bitcoin in the world. It is the price of the small, restless sliver that happens to be willing to trade today – the bottom rung. To buy more, you have to climb.
The margin
Why heavy buying can coincide with a falling price
Here is the part that feels like a paradox and isn't. In the short run, price can fall while serious money is accumulating hard. The two facts live in different layers of the market.
Price is set at the margin – by whoever is willing to sell into the bid right now, not by the silent majority who would not part with a coin under a million dollars. When a long-dormant whale finally distributes, or an early holder from a decade ago quietly sells a tranche into the new institutional demand – a kind of silent IPO, an offering with no prospectus – those coins meet the buyers and the clearing price can sag, even as the buyers are vacuuming up everything they can. Add leveraged traders forced out of positions, and the housekeeping of ETF authorized participants shuffling coins in and out to keep funds balanced, and you get plenty of short-run selling pressure layered on top of relentless accumulation.
So accumulation does not announce itself in the price. It shows up first in who holds the coins – in the slow migration of supply from restless hands toward patient ones, higher up the ladder. The price chart is the noise. The ownership chart is the signal. And here the direction matters more than any single figure. On-chain analytics firms that track wallet behaviour – Glassnode and CryptoQuant among them – report that in the tightest stretches, long-term holders and large wallets absorb coins at several times the rate new ones are mined; one rough read puts it around three times annual issuance at the extremes.1 Treat that number as illustrative and volatile – it swings hard, and methodologies differ. The point the essay leans on is not the multiple but the sign: through these spells, coins leave the willing float faster than fresh supply replaces them. None of that is visible in any single day's red candle.
One honest caveat about the usual evidence for this. People often point out that some seventy percent of all bitcoin "hasn't moved in over six months," and treat that as proof the coins are unavailable. It is not. A coin can be sold in an instant without having moved recently; dormancy measures the absence of recent transactions, not a vow of silence. What it really is is a proxy for conviction – a reasonable read on how many holders are sitting tight – not a guarantee that any given coin is off the table. The honest claim is softer and still damning: the longer a coin sits, statistically the higher up the reservation ladder its owner tends to be, and the more it would cost to coax it loose.
A coin can be sold in an instant
without ever having moved.
The non-linearity
A fixed supply does not re-price gradually
Now run the migration forward. Committed buyers keep absorbing coins: corporate treasuries (some locking them up to earn yield through instruments like STRC), spot ETFs already holding around one and a half million coins,2 sovereigns testing the water, and individuals who simply will not sell below seven figures. Every one of them takes float off the table – permanently.
For a while, nothing dramatic happens to the price. The bottom rung still has coins on it; the squeeze is invisible because the cheap sliver hasn't run dry. But that sliver is finite, and it is shrinking. The mechanism is not that sellers vanish – it is that the price required to coax out the next coin rises faster and faster as the willing float thins. And this is the part to be careful about: it is not an empirical claim that needs the on-chain numbers to hold up. It is a deduction, and it follows from just two premises, both beyond dispute. One: the supply is fixed, so quantity cannot expand to meet demand. Two: holders do not share a single selling price – their reservation prices are heterogeneous, some willing at a small premium, others only at a multiple, a stubborn tail at no number they expect to live to see. Grant only those two and the rest is forced. Sort the holders low to high and you have a ladder. Now spend into it. The cheap supply clears first – by definition, because it is the cheapest, so it goes first – and once it is gone, the only holders left are the ones who always wanted more. The buyer is therefore forced strictly upward: clear the holders who would sell at this price and the next coin sits with someone who wanted half again as much; clear them and the one after wanted double. The ladder does not merely rise, it steepens, because the survivors are self-selected for stubbornness – every coin you buy removes a willing seller and leaves behind a less willing one. So the marginal price cannot climb in even steps; it must accelerate. The on-chain statistics later in this piece are evidence that this is happening now, but they are not what makes it true. Heterogeneity plus a fixed supply makes it true, the way two and two make four.
This is the difference between a market where supply is a knob you can turn and a market where supply is a steepening wall. Against a knob, demand pushes and the knob gives a little. Against this wall, demand pushes and for a long while the price barely yields – and then the rungs above thin out, the marginal price lurches, and the entire adjustment that should have been spread over years arrives in a span you can measure in days. Quantity can't expand to absorb the demand, so price does the absorbing – and it does it non-linearly. The widget below lets you feel the shape of it.
When quantity can't move,
price is the only thing left that can.
The projector
Watch the liquid float drain
A toy model, not a forecast. It tracks one thing: the pool of coins on offer near today's price – the bottom rung. Each week the committed buyers take their share, a trickle of newly mined coins refills it, and the pool moves toward the pressure point: the week when that cheap sliver thins out and the next coin can only be bought far higher up the ladder.
Drag the inputs – the squeeze in motion
A toy model. Net weekly drain = buying − issuance, held constant; real demand, supply and price all vary. Crucially, a real market re-prices well before the float literally hits zero – rising price coaxes some holders to sell and cools some buying. Read the crossing not as a date but as a pressure point: where quantity can no longer do the adjusting, and price must.
One honest caveat about what the widget actually draws. It shows the float thinning, not the price path. The line you scrub is the depletion of the cheap supply – the cause; the non-linear price lurch this essay describes is the reflexive consequence of that depletion, which the toy model deliberately does not attempt to plot. Read the crossing as a pressure point, not a price target. With that line held straight, the lesson the model makes physical is in the gap between the two: hold the buying steady and the cheap float falls in a near-straight line – but the consequence for price is anything but straight. For most of the run, nothing visible happens. The drama is all stored up, invisible, until the bottom rung thins out and the adjustment that quantity refused to make lands entirely on price, which climbs the steepening ladder all at once.
The lesson
Falling price, tightening grip
Strip away the day-to-day and a flat principle remains. A red candle tells you what the marginal seller did this afternoon. It tells you almost nothing about the squeeze.
If most holders treat bitcoin the way the readers of this site tend to – as a lifeline meant to outlast them, not to be sold until seven figures and even then only sparingly, often locked up to earn yield rather than spent – then each conviction that hardens lifts another coin further up the reservation ladder. The cheap rung gets shallower with every cycle. A falling price during heavy accumulation is not evidence against the thesis; under a fixed supply it is exactly what the early innings of a squeeze look like. The mistake is to read a smooth-line story onto a market whose defining feature is that it cannot stay smooth.
Now the objection a careful skeptic actually presses, and it is a strong one: every prior squeeze was followed by a brutal drawdown – often seventy to eighty percent. The wall did not just move up; it then moved sharply down. If the float can re-fatten that violently, what is structural about any of this? The honest answer is to separate two things the price chart smears together. The cyclical part is real: most of those crashes were leveraged positions and speculative late-comers being flushed out, paper hands that had bought the top on borrowed money. That flush is genuine and recurring, and nothing here denies it. But underneath the flush runs a slower current – the structural one. Through each boom and bust, the same on-chain trackers find the share of supply held in long-term, low-activity wallets trending upward across cycles, and the coins available near any given price trending down.1 Read those metrics as approximate and contested – they rest on heuristics about which wallets are "long-term," not a census – but the trend across cycles is the durable part, not any one reading. The drawdowns are the cyclical wall breathing out; the rising illiquid base is the structural wall ratcheting in. A squeeze does not promise the price only goes up. It promises that the floor under it keeps rising, because each cycle tends to leave a little more of the supply in the hands that held through the crash than was there before it.
The drawdowns are the wall breathing out.
The rising illiquid base is it ratcheting in.
And here is the non-obvious capstone the deduction has been building toward, the one worth carrying away: a squeeze needs no conductor, and that is precisely what makes it impossible to stop. Look again at what the mechanism actually required. Nowhere did it need anyone to coordinate. It ran on heterogeneous reservation prices and a fixed supply – on each holder privately naming a number and the ladder steepening as the cheap rungs cleared. That is the deep reason it cannot be switched off. The instinct is to hunt for an orchestrator – a cartel, a whale, a plan – and to assume that if no one is conducting, nothing is happening. But this is the rare engine where the absence of coordination is the source of its force. A patient accumulator might ride it deliberately, sure; it does not require one. When enough holders, each for their own private reasons, independently decline to sell at today's price, the arithmetic does the rest – no participant chooses the break, and so no authority can permit it away, because there is no switch to flip and no committee to lean on. You cannot subpoena a Schelling point. A coordinated corner can be broken by breaking the coordinator; an uncoordinated one has no head to cut off. That is what the heterogeneity premise quietly bought us at the end: a result with no author, and therefore no one to stop. That same uncoordinated conviction is the whole subject of the quiet believers, worth reading on its own terms. What matters here is only the mechanism: the coins are not cheap at today's price, and a market that has to climb its own ladder to find the next seller has only one way to resolve.
Demand pushing on an expandable supply nudges the price. Demand pushing on a wall does nothing – and then, all at once, does everything. The whole argument is in that difference, and the only honest forecast is its shape, not its date.
Still skeptical
If a red candle still looks like the thesis breaking, read why the swings are the toll, not the trip – and the case in full.
Volatility Is the Toll, Not the Trip → The Asset No Empire Can Freeze →Curious
Why would so many holders independently refuse to sell? The incentives, and the asymmetry that makes holding rational.
The Game Theory of Bitcoin → Take the Zero Off the Table →Convinced
You see the wall. Now follow where the standing bids come from – the index trap and the treasury wrapper.
The Index Trap → Saylor's iPhone Moment →Sources & notes. The on-chain figures below come from third-party analytics firms (illiquid-supply and long-term-holder metrics of the kind published by Glassnode and CryptoQuant-style trackers); they rest on wallet heuristics, not a census, move constantly, and differ by methodology. The argument deliberately leans on the direction of these series, not the precise level. 1. In recent stretches, long-term holders and large wallets have absorbed coins on the order of several times new issuance – one rough read is ~300% of annual issuance in the tightest spells (an illustrative on-chain estimate that swings widely). Supporting figures, all approximate and moving: illiquid supply on the order of ~70–75% of circulating supply; long-term-held supply a similar majority; exchange balances down sharply (roughly halved) since late 2024. These are estimates from on-chain analytics, not audited counts. The ~70–80% peak-to-trough drawdowns are a recurring feature of prior cycles; across them the long-term-held share of supply has trended up. Dormancy ("hasn't moved in 6+ months") is a proxy for holder conviction, not proof a coin is unavailable. 2. Spot ETFs holding ~1.5M BTC; new issuance ~3,150 BTC/week post-2024 halving – both approximate and drifting. Not investment advice – this piece explains a market structure, not a trade.