Bitcoin’s Stock-to-Flow Model: Why S2F Keeps Missing the Price
Bitcoin slid under $59,000 in late June 2026, far from the $288,000 the Stock-to-Flow model once implied. We explain why statisticians and economists say S2F was never a real forecast.
Bitcoin spent the final week of June 2026 sliding toward price levels that its most famous valuation model once implied it would leave behind for good. By the morning of June 26, a single coin changed hands near $58,980, down about $2,300 on the day and far below the roughly $72,800 it had touched earlier in the month, according to Fortune’s daily Bitcoin price tracker. For investors who built a thesis on the Stock-to-Flow model, that figure is uncomfortable. The model, known to traders as S2F, once pointed to a Bitcoin price of $288,000 for the current four-year cycle and was treated for years as near proof that scarcity alone would carry the asset toward seven figures. The 2026 downturn is only the newest reason to revisit a question that statisticians raised almost from the start: was S2F ever a forecast, or just a chart that happened to look persuasive? Six years of out-of-sample results now make that easier to judge.
What the Stock-to-Flow Model Actually Says
Stock-to-flow is a scarcity ratio borrowed from the world of commodities. The stock is the amount of an asset already in existence, and the flow is how much fresh supply arrives each year. Gold scores high, with a ratio above 50, because the above-ground stockpile dwarfs annual mine output, which is part of why it has held monetary value for centuries. Bitcoin’s flow shrinks roughly every four years at each halving, when the reward paid to miners is cut in half, so its S2F ratio rises in steps and now sits in the same range as gold. The pseudonymous analyst PlanB plotted that rising ratio against Bitcoin’s market value and fit a straight line through the dots. In his original write-up he reported a 95 percent R-squared and argued that scarcity, as measured by the ratio, directly drives value. Put plainly, the claim was that you could read Bitcoin’s future price straight off its issuance schedule.
From a 2019 Blog Post to a Movement
That write-up, titled “Modeling Bitcoin Value with Scarcity,” appeared on March 22, 2019, from an author who described himself as a former institutional investor with 25 years in financial markets. The piece leaned on older ideas, including Nick Szabo’s notion of unforgeable costliness and Saifedean Ammous’s writing on hard money, then wrapped them in a single regression line that was easy to share. That simplicity was the appeal. A logarithmic chart with price riding a tidy upward channel gave the “number go up” crowd something that looked rigorous, and it spread quickly across crypto social media. Within a year the S2F line had become a cultural touchstone, reproduced on dashboards, podcasts, and trading-desk slide decks as if it were a law of nature rather than a blog post. For a long stretch it served as the default bull case, repeated confidently by people who had never inspected the underlying regression.
The Forecasts That Defined the Model
In April 2020, PlanB extended the idea with a cross-asset version called S2FX, which folded gold and silver into the same regression and produced a headline target: an average Bitcoin price of $288,000 across the 2020 to 2024 cycle, implying a market value near $5.5 trillion. That single number became one of the most quoted price targets in the industry, plastered across forums and pitch decks. PlanB’s follower count climbed into the hundreds of thousands, and several mainstream finance outlets began citing the chart as if it were a settled valuation method rather than one analyst’s hypothesis. A year later he layered on a separate “floor model” that issued specific month-end price floors for the back half of 2021. Those figures, starting at $47,000 in August and climbing to $135,000 in December, turned a long-run scarcity thesis into a short-term prediction that anyone could check against the tape within weeks. That precision cut both ways. It made the model feel testable and serious, and it also handed critics a scoreboard.
When the Floor Model Met Reality
The floor model looked uncanny for three months, then broke. August, September, and October 2021 each closed near their targets, which PlanB cited as confirmation. November and December did not come close. By late November he conceded that the $98,000 target would probably be the floor model’s first miss. December’s outcome was the starkest: against a $135,000 target, Bitcoin finished the year near $46,000, a gap of nearly $90,000 that no amount of framing could explain away. The table below sets each monthly target against Bitcoin’s approximate month-end closing price, drawn from historical data compiled by CoinGecko.
| Month (2021) | Floor model target | Approx. month-end close | Result |
|---|---|---|---|
| August | $47,000 | $47,100 | Met |
| September | $43,000 | $43,800 | Met |
| October | $63,000 | $61,300 | Met |
| November | $98,000 | $57,000 | Missed |
| December | $135,000 | $46,200 | Missed |
The Statistical Case Against S2F
The deeper problem is not one bad quarter; it is how the model is built. Writing in CoinDesk, Nico Cordeiro of the research firm Strix Leviathan argued that fitting a regression to two series that both trend upward over time is a classic recipe for spurious correlation, where the math reports a tight relationship that carries no predictive power. In statistics this is the well-known trap of non-stationary data: line up almost any two quantities that rose through the 2010s and they will look correlated. He also flagged a circularity. Bitcoin’s market value equals price multiplied by stock, while stock sits inside the S2F ratio on the other side of the equation, so the regression partly explains stock with itself. Because the ratio barely moves from month to month, Cordeiro noted, the fit mostly extrapolates past growth forward, which is how the same math can imply a single Bitcoin worth $235 billion by 2045. Defenders countered that the two series are cointegrated, a property that can make a long-run link genuine, but independent statisticians who reran the tests found the result too fragile to lean on. Bitcoin Magazine reached a similar verdict, calling the model statistically hollow once the shared trend is stripped out.
Scarcity Is Only Half of a Price
Even setting the statistics aside, the model rests on a strong economic claim: that supply scarcity, on its own, sets value. A critique published by the Mises Institute points out that price is set where supply meets demand, and S2F carries no demand term at all. Plenty of things are extremely scarce and close to worthless, from one-off concert posters to obscure collectibles, simply because few people want them. Gold’s price is not dictated by its stock-to-flow ratio alone; it moves with real yields, central-bank buying, and sentiment. A framework that ignores demand cannot explain why Bitcoin can fall 70 percent while its issuance schedule, and therefore its S2F ratio, barely changes. Demand can evaporate in a week, while the supply schedule cannot speed up to compensate, nor slow down to defend a price. The 2026 selloff is a tidy example: supply did exactly what the protocol promised, and the price dropped anyway.
2022 and the Public Reckoning
By the middle of 2022 the gap between model and market had become a spectacle. Ethereum co-founder Vitalik Buterin argued that price models projecting a false sense of certainty deserve the mockery they get, posting his critique with Bitcoin trading near $21,000, well below the six-figure zone the model implied for that year. Worse followed. When the FTX exchange collapsed that November, Bitcoin slid below $16,000 to a two-year low, roughly a sixth of the price S2F implied for the cycle. PlanB acknowledged at one stage that the model looked invalidated, then later maintained the longer-run thesis was still intact, a back-and-forth that did little to settle the argument. For many newer investors, that drawdown was a first hard lesson that a confident-looking model and a profitable trade are not the same thing.
Where the 2026 Slump Leaves the Model
The April 2024 halving cut the block reward to 3.125 BTC and pushed Bitcoin’s S2F ratio higher again, the kind of step the model says should precede another surge. Instead, after setting records in 2025, Bitcoin rolled over through the first half of 2026 and traded under $59,000 in late June. The arrival of spot ETFs and heavy institutional flows has also made the asset trade more like a macro instrument than a pure scarcity play. Some S2F supporters have responded by stretching their targets toward $500,000 and beyond, which is exactly the behavior Buterin warned about. A model that can absorb any outcome and still call it confirmation is not really making a prediction. When every price, up or down, can be folded back into the thesis, the thesis stops carrying information, and it certainly stops being a sound basis for risking money.
What Serious Analysts Watch Instead
Desks that once kept an S2F chart open have mostly moved to measures that include demand. On-chain metrics such as realized capitalization, MVRV, and spent-output ratios try to gauge what holders actually paid and whether they are sitting in profit or loss. Since US spot Bitcoin exchange-traded funds launched in January 2024, daily creation and redemption flows have become a closely watched proxy for institutional appetite. Macro inputs matter too. The June 2026 drop has been tied to sticky inflation, fading hopes for Federal Reserve rate cuts, a firmer dollar, and accelerating ETF outflows, none of which appear anywhere in a stock-to-flow ratio. None of these tools is a crystal ball either, but each at least lets demand into the picture. The broader shift is away from any single master equation and toward a dashboard of signals, on the view that no one number can capture a market driven by liquidity, regulation, and human conviction at the same time. That blind spot is precisely why scarcity-only models keep getting caught flat-footed when sentiment turns.
What This Means for US Investors
The lesson is not that Bitcoin is doomed; it is that a confident-looking chart is not a guarantee. The SEC’s Office of Investor Education and Advocacy urges caution with crypto asset securities and tells investors to question any pitch that promises a fixed number or a sure thing. S2F was never sold as fraud, and PlanB has always framed it as a model rather than a promise, but its career is still a clean study in model risk: a single equation, fit to a short and unusual price history, then dressed up as destiny. The model did not fail because Bitcoin is worthless; it failed because a tidy line through past prices was asked to do work that no single variable can. Treat price models as one input among many, size positions for the real chance that they are wrong, and remember that a forecast which cannot fail is not really a forecast.
By the HOGE Wire editorial desk, covering crypto markets, the models that try to price them, and the people who bet on both.