Liquidity Beat Scarcity: Bitcoin’s Stock-to-Flow Model in 2026

The model that got the supply right and the price wrong

The stock-to-flow model made two promises. One was easy to check and one was not. The easy promise was about supply: Bitcoin’s issuance would keep halving on a fixed schedule, scarcity would deepen, and the stock-to-flow ratio would climb. The hard promise was about price, that deepening scarcity would drag the market up a predictable curve toward six figures and then seven. In 2026, the two promises finally came apart in public. The supply behaved exactly as coded. The price did not.

That split is the whole story, and it is why a model that looked almost prophetic in 2020 now reads like a cautionary tale. Bitcoin traded near $81,400 on 20 October 2026, roughly 35% below its October 2025 record of $126,080, according to CoinGecko. The April 2024 halving arrived on time and cut new issuance in half, exactly as stock-to-flow assumes. Scarcity rose. Price went sideways, then down. When a framework’s single most reliable input produces the wrong output, the framework is not measuring what it claims to measure.

This is not another round of mocking a pseudonymous analyst. It is an argument about what actually sets Bitcoin’s price in 2026, and why the honest answer is liquidity rather than scarcity. The halving is real. So was the $19 billion of leverage that evaporated in a single October 2025 session, the largest liquidation day on record, according to CoinShares. One of those events lives inside the stock-to-flow model. The other, the one that actually moved the tape, is invisible to it. The gap between those two facts is the subject of this piece, and it matters for anyone still reading the halving calendar as if it were a price forecast.

What stock-to-flow actually claims

Stock-to-flow is a scarcity ratio borrowed from commodities markets. Stock is the total existing supply; flow is annual new production. Divide stock by flow and you get the number of years of current output it would take to recreate the entire existing supply. Gold sits around 62 on this measure and silver around 22, according to CoinGecko’s explainer. The higher the ratio, the harder the asset is to inflate, and the thinking goes that harder money should command a higher value over time.

In March 2019, the pseudonymous analyst known as PlanB published Modeling Bitcoin’s Value with Scarcity, regressing Bitcoin’s market value against its stock-to-flow ratio and reporting an R-squared near 95%. Bitcoin traded around $4,002 at the time. The appeal was mechanical clarity: each halving cuts flow, roughly doubling the ratio, and the regression line converts that into a far higher price. Bitcoin’s ratio sat near 27 before the 2020 halving and climbed to roughly 119 after 2024. For a stretch of 2020 and 2021, the chart it produced was the single most shared image in Bitcoin, a straight diagonal that seemed to promise the destination in advance.

A year later, PlanB extended the idea into a cross-asset version called S2FX, which placed Bitcoin on a single line with gold and silver and produced an average target near $288,000 for the 2020 to 2024 cycle, with a reported R-squared of 99.7%, as set out in PlanB’s own write-up. The current headline figure is a $500,000 average for the 2024 to 2028 cycle, inside a $250,000 to $1 million band, a number PlanB still reiterates, as Yahoo Finance has reported. The table below shows how the ladder is supposed to climb.

The one variable stock-to-flow gets right

Give the model its due, because the concession is what makes the rest of the argument airtight. Bitcoin’s supply schedule is not a forecast; it is code. The block subsidy began at 50 BTC, halved to 25, then 12.5, then 6.25, and dropped to 3.125 BTC at the April 2024 halving. The next halving, expected around April 2028, will cut it to 1.5625 BTC, per CoinGecko’s halving tracker. Every node on the network enforces this rule. There is no committee, no discretion, and no surprise issuance.

So of the two inputs stock-to-flow needs, one is essentially perfect. Scarcity is knowable years in advance, down to the individual block. That is precisely why 2026 is such a clean test. When a model with a flawless input still produces a wildly wrong output, you cannot blame the data. You have to examine the theory that connects the input to the output. The theory is that supply-driven scarcity sets the price. That theory is what broke, and it broke in the open, at the one moment the model was supposed to be at its strongest.

The cleanest falsification: the halving happened, the price did not follow

State the test as plainly as possible. The 2024 halving occurred on schedule. Flow was cut in half. The stock-to-flow ratio roughly doubled. By the model’s own logic, the eighteen months that followed should have carried Bitcoin into the $288,000 to $500,000 zone. Instead, Bitcoin peaked near $126,080 in October 2025, a cycle gain closer to 2x than to the 5x or 10x the model implied, then drifted lower across the next year to the low $80,000s. Scarcity deepened exactly as predicted, and the price went the other way.

This is not a near miss that better calibration would fix. It is the central claim failing at the precise moment it should have been strongest. The pattern is not new, either. In 2021 the model’s floor variant called for a $98,000 November close that never arrived; PlanB posted that it was a “FLOOR MODEL FAIL” while insisting the broader model remained intact, as CryptoPotato documented. Bitcoin then slid through the model’s lower band in mid-2022 and kept sinking through the FTX collapse that November, far beneath any price the scarcity line allowed. Five years on, the behavior is unchanged: the supply forecast lands, the price forecast does not, and each failure is absorbed rather than counted.

2026 was a liquidity market, not a scarcity market

If scarcity did not move Bitcoin this cycle, what did? The honest answer is a stack of demand and liquidity variables, none of which appears anywhere in the stock-to-flow equation. Start with demand. In January 2024, the SEC approved eleven spot Bitcoin ETFs, opening a regulated pipe between Wall Street balance sheets and the asset. That single structural change rerouted how marginal demand reaches Bitcoin, and it has nothing to do with how many coins miners mint per day.

Then layer on the rest. Corporate and sovereign treasuries added a new structural bid in the bull phase and became forced sellers in stress. The spread of Wall Street products put crypto beside equities on the same brokerage screens, binding Bitcoin tighter to risk appetite. Federal Reserve policy set the discount rate. Global dollar liquidity set the tide. And leverage, the most violent input of all, set the short-term swings. Stock-to-flow sees none of it. It sees the issuance schedule and nothing else. The table below lists the drivers that actually set Bitcoin’s price across 2024 to 2026 and marks the one, and only one, that the model can detect.

Put bluntly, a model that captures one of six price drivers, and specifically the driver that did not change Bitcoin’s trajectory this cycle, is not a price model. It is a supply calendar with a regression line bolted on.

Flows made the point daily. In the stock-to-flow worldview, the coins mined per day are the only supply story that matters, and after the 2024 halving that figure fell to roughly 450 BTC a day (about 144 blocks at 3.125 BTC each). Yet the spot ETFs alone have absorbed many multiples of daily issuance on heavy inflow days and disgorged it on heavy outflow days, which means the marginal coin that sets the price is changing hands for reasons the issuance schedule cannot describe. When ETF demand runs hot, price rises into a fixed supply; when it reverses, price falls into the same fixed supply. Supply is the constant and demand is the variable, so a model built on the constant was always going to explain the least about where the price actually went.

The Bessent buyback: the shock a supply model cannot see

The clearest illustration came in August 2026. On 19 August, Treasury Secretary Scott Bessent doubled the maximum size of long-dated bond buybacks to at least $4 billion per operation, up from $2 billion, across the 10, 20, and 30-year maturities. The stated aim was to pull down stubborn long-term yields. It did not work on yields, which stayed near 5.25% on the 30-year. It worked on Bitcoin, which surged toward $80,000 within hours and triggered billions of dollars in short liquidations, as CoinDesk reported under a headline noting the buyback wanted lower yields and got a Bitcoin surge instead.

Nothing about Bitcoin’s supply changed on 19 August. No coins were burned or minted off schedule. The stock-to-flow ratio that afternoon was identical to the morning. Yet the price jumped, because traders read the buyback as a signal that officials would lean toward easier liquidity, a kind of stealth easing. A supply-only model is structurally blind to the single biggest same-day catalyst of the late-summer tape, and that blindness is not a bug a tweak can patch. It is built into the premise, which holds that supply is the thing that matters and demand will sort itself out.

Zoom out and the August episode fits a bigger pattern that scarcity models have no vocabulary for. With deficits wide and long-term yields sticky, markets increasingly treat Treasury operations and the Federal Reserve’s balance sheet as the real drivers of asset prices, a regime some analysts call fiscal dominance. In that world, Bitcoin trades less like digital gold with a fixed mint schedule and more like the longest-duration risk asset on the board, rallying when liquidity looks set to expand and selling off when it tightens. None of that is visible to a ratio of stock over flow, which is why the model had nothing to say about the single most important macro signal of the late summer.

The anti-stock-to-flow target: Connors and the liquidity thesis

The liquidity thesis has its own price targets, and they are built from the opposite inputs. Mark Connors, chief investment officer at Risk Dimensions, frames $180,000 as Bitcoin’s first meaningful price threshold and sketches a cycle range of $180,000 to $360,000 through 2030, driven not by scarcity but by the expansion of dollar liquidity. “When that happens, that’s when bitcoin starts to seek that first $180,000 price threshold,” he told CoinDesk, pointing to Treasury buyback support he believes could eventually reach $10 billion to $30 billion a month.

The contrast is the useful part. A stock-to-flow target answers the question of how scarce Bitcoin is. A liquidity target answers how much money is moving through the system looking for a home. In 2026, the second question tracked the price and the first did not. Even the more conservative institutional calls are built this way: Citigroup’s 12-month target of $113,000, cited by coin-turk, rests on flows and macro conditions, not on the issuance calendar. Nobody setting a serious price target in 2026 is drawing a line off the halving. They are all, bulls and bears alike, arguing about demand.

October 2026: the tape proves the point

You do not need to reach back to August for the evidence; the October 2026 tape makes the case in real time. US spot Bitcoin ETFs have leaked money for much of the month, with net outflows of roughly $451 million over seven trading sessions and a single session on 16 October shedding $82.4 million, according to coin-turk. The same desks that set near-term ranges are anchoring them to liquidity rather than supply: QCP Capital told clients Bitcoin was likely to trade “within a range of $80,000 to $90,000 unless there was a clear improvement in market liquidity.”

Read that projection again. A trading desk setting a Bitcoin range in October 2026 anchored it to liquidity, not to the stock-to-flow ratio, which has only risen since the halving. Meanwhile, the market’s attention is fixed on the Federal Reserve’s 27 to 28 October meeting. After September’s hot 3.7% CPI print, traders still price roughly an 85% chance of a hold, with a hike seen as more likely at the December meeting, according to Forbes. The Fed had already lifted its target range to 3.75% to 4.00% in September, its first hike since 2023. Bitcoin’s bounce on 19 October, coin-turk noted, tracked fading expectations of an imminent hike.

The price, in other words, is pinned to a rate decision and a liquidity backdrop, not to a supply calendar. The same prediction-market machinery that now sets the odds on elections and policy is where traders read the Fed, and it is those odds, not the halving, that move Bitcoin from week to week. This is the regime our year-end outlook described as a market stuck between the Fed and the low $80,000s. Scarcity has not changed in any of it. Liquidity has changed constantly, and so has the price.

The statistics never supported scarcity as destiny

Even before 2026 delivered its real-world verdict, the statistics never supported scarcity as destiny. The most cited teardown, written by the analyst Level39 for Bitcoin Magazine, lays out three problems. The first is a tautology: market value equals stock times price, so placing market value on one axis and a stock-based ratio on the other means stock is partly explained by stock. The second is autocorrelation: each day’s value is largely yesterday’s value, so the data points are not independent observations. The third is spurious regression: fit two upward-sloping lines against each other without detrending and you get a beautiful correlation that signifies nothing.

Cory Klippsten, chief executive of Swan Bitcoin, put it bluntly in that analysis: “The model is autocorrelated. Does not need S2F at all. Just price vs time.” Adjust for the autocorrelation, the same piece notes, and the R-squared that impressed everyone collapses toward zero. There is also the matter of sample size: the entire model rests on four halving epochs, which is nowhere near enough observations to establish a law of nature. Vitalik Buterin was characteristically direct in June 2022, writing that “financial models that give people a false sense of certainty and predestination that number-will-go-up are harmful and deserve all the mockery they get,” as The Block reported the day Bitcoin first broke through the model’s lower band.

The chameleon that refits every time it misses

A model that keeps missing but keeps getting re-fitted is not being tested; it is being redecorated. The original 2019 model implied roughly $55,000 as a cycle average. The 2020 cross-asset revision raised that to $288,000. The current figure is $500,000. Each miss was followed by a recalibration to a higher, later target, which is how a framework stays alive without ever being wrong in a way that costs its author anything.

Nico Cordeiro of the research firm Strix Leviathan named this dynamic years ago, calling stock-to-flow a “chameleon model” and arguing it was “about as successful at forecasting bitcoin’s price as the astrological models of the past were at predicting the future,” in a 2020 CoinDesk critique. A target with no deadline, or one that resets each time it is breached, cannot be falsified, and a claim that cannot be falsified is not a forecast, however precise its dollar figure looks. The scorecard below collects the model’s named calls against what the market actually did.

Defenders answer that the model is about direction over whole cycles, not month-to-month precision, and that believers were simply early. But being early is not the same as being right. A forecast with no deadline cannot be tested, and a target that keeps sliding into the future absorbs every miss as a delay rather than a refutation. By late 2026 the honest reading is not that $500,000 is merely running behind schedule; it is that the mechanism the number depends on, scarcity dragging price up a fixed curve, did not operate this cycle at all. A clock that is always about to be right is not a clock.

Scarcity lens versus liquidity lens

It helps to put the two frameworks side by side, because the point is not that stock-to-flow is stupid; it is that it answers a different question than the one investors actually face. Scarcity is a property of the asset. Liquidity is a property of the environment the asset trades in. For most of Bitcoin’s history, the two were correlated enough that you could mistake one for the other. In 2026, with the supply schedule fully priced in and the macro backdrop swinging hard, they decoupled, and only one of them tracked the tape.

The liquidity lens is not a crystal ball. It can be wrong, and liquidity can reverse without warning, as the October outflows show. But it has one property stock-to-flow lacks: it is measurable in close to real time, through ETF flows, net liquidity, funding rates, and the Fed’s own balance sheet. That makes it testable, and a framework you can test is worth more than one you can only believe.

The power law and the four-year cycle inherit the same flaw

When stock-to-flow became indefensible, its following did not disband; it migrated to successor models, chief among them the power law, which fits price to time since the genesis block raised to a fixed exponent. The power law has aged better than stock-to-flow, but it fails the test that matters: it still models price as a function of time, not of demand. Klippsten’s “just price vs time” critique cuts against it as cleanly as against S2F, and its own practitioners concede it does not mechanistically incorporate ETF demand or liquidity cycles. Better geometry, same determinism.

The four-year cycle, the other great determinism, met the same fate. Matt Hougan, chief investment officer at Bitwise, told clients in a December 2025 memo that “the four-year cycle is dead” and that Bitcoin had entered a “ten-year grind” shaped by adoption and macro rather than halving clockwork, as Bitwise published. Each successive cycle has also peaked at a smaller multiple of the previous high, a steady deceleration rather than the ever-larger moves that scarcity determinism implies. There is a tidy explanation for the one time stock-to-flow appeared to work. The 2020 to 2021 run is better read as reflexivity than as scarcity: the model became a famous chart, traders coordinated on its targets, and for a while the forecast helped manufacture the rally it predicted. That Schelling-point effect needs a crowd of retail believers aiming at the same number. In 2026 the marginal buyer is an ETF desk, not a forum, and the self-fulfilling loop is gone.

Even the model’s author navigates by other instruments

The most telling evidence that scarcity lost its crown is where PlanB himself looks for signals in 2026. He still points at the $500,000 stock-to-flow number as a long-run value signal, but his day-to-day commentary now leans on 200-week and other moving averages, realized price, the relative strength index, and mining economics. In September 2026, he told readers that mining difficulty “better reflects market price than models such as S2F or power laws,” per KuCoin.

That is remarkable. The creator of the scarcity model navigates the real market with price-versus-time technicals and cost-of-production data, exactly the demand and cost inputs that stock-to-flow was supposed to render unnecessary. When the author keeps the famous number on the wall but steers by other instruments, the number has become a narrative, not a tool. There is no shame in that. A good narrative can still teach something true about Bitcoin’s long arc, about why fixed supply matters over decades. It just cannot tell you where the price will be next quarter, which is the one thing the chart was sold as being able to do.

How to read any Bitcoin price model in 2026

None of this means price models are useless. It means they should be read as what they are. A model that takes only time or only supply as an input can describe the past beautifully and still tell you nothing new, because it has no mechanism to react to the things that actually move the market. Ask any model four questions before you trust it.

• Does it make a specific, dated, falsifiable prediction, or a target that quietly slides forward?
• Does it model demand, or only the passage of time and the supply schedule?
• Has it ever issued a sell signal, or is it permanently bullish by construction?
• Are its bands tight enough to size a position, or so wide that any outcome confirms it?

If the honest answer to the third question is always up, you are holding a belief, not a forecast. Regulators make a related point in plainer language. When the SEC approved spot Bitcoin ETFs, Chair Gary Gensler stressed that the agency was not endorsing Bitcoin and that investors should stay cautious about a volatile, speculative asset, in his January 2024 statement. Past performance, including a backtest with a 95% R-squared, does not guarantee future results.

The on-chain and macro indicators worth watching today, realized price, MVRV, net liquidity, and funding rates, describe conditions rather than promise destinations, and that humility is a feature, not a weakness. The lesson of 2026 is narrow and durable: in a market where one liquidity headline can add or erase tens of thousands of dollars of value in an afternoon, a model that cannot see liquidity was always going to miss. Scarcity is still real, still fixed, and still worth understanding. It is simply not the variable setting the price right now, and pretending otherwise has cost people real money. The sooner that distinction is clear, the easier it becomes to read every chart, target, and rainbow band that scrolls across your feed.

Frequently Asked Questions

What is the Bitcoin stock-to-flow (S2F) model?

It is a scarcity-based valuation model that divides Bitcoin’s existing supply (stock) by annual new issuance (flow) and uses that ratio to project price. Introduced by the pseudonymous analyst PlanB in 2019, it reported a 95% R-squared fit and implied that each halving, by cutting issuance, should push Bitcoin to progressively higher prices.

Why did the stock-to-flow model fail in 2026?

The 2024 halving deepened scarcity exactly as the model assumed, yet Bitcoin peaked near $126,080 and then fell toward $81,000, nowhere near the model’s $500,000 cycle target. Scarcity was not the variable setting the price; liquidity and demand such as ETF flows, Fed policy, and Treasury operations were, and stock-to-flow cannot see any of them.

What is PlanB’s current Bitcoin price prediction?

PlanB still cites a $500,000 average for the 2024 to 2028 cycle, within a $250,000 to $1 million band. In practice he now leans on moving averages, realized price, the RSI, and mining difficulty, which he has said reflect the market better than stock-to-flow or power-law models.

Is scarcity irrelevant to Bitcoin’s price?

Scarcity is real and fixed, but by 2026 it had stopped being the marginal driver of price. Because the supply schedule is fully known in advance, it is already priced in; the swing factor is demand and liquidity, which change daily and are what the market actually reacts to.

What are better alternatives to the stock-to-flow model?

There is no reliable crystal ball, but indicators that describe conditions rather than promise targets tend to be more honest, including realized price, MVRV, net liquidity, and funding rates. Liquidity-based frameworks, such as Mark Connors’ $180,000 to $360,000 range, at least track the variable that actually moved Bitcoin in 2026.

Priya Reddy covers Bitcoin markets, on-chain data, and price models for HOGE Wire.