Stock-to-Flow on Trial: Bitcoin a Year After the Top

A year ago this month, Bitcoin printed the highest price in its history, $126,198 on 6 October 2025, according to figures compiled by Coinlaw. Four days later it suffered the worst single-day wipeout the market had ever seen, a cascade that CoinShares pegged at more than $19 billion in liquidated positions. This week, Bitcoin trades near $82,000, down roughly 32% over the year.

For the stock-to-flow model, the anniversary of that top is less a birthday than a day in court. By the slope that made S2F the most-cited Bitcoin valuation chart of the last cycle, the coin should be deep into six figures by now and climbing toward seven. It is not. The fairest way to judge a forecast is to look at it when it mattered most, at the peak, and a full year past that peak the verdict is unflattering. This is the accounting of how the most famous Bitcoin price model ever drawn got built, why it hypnotized the market, and why the statistics underneath it never held.

A Model on Trial at the Top

Most price predictions are safe because they never come due. Stock-to-flow was different. Its author, a pseudonymous analyst who writes under the handle PlanB, pinned numbers to calendar dates, and in one case staked his credibility on a hard deadline. In June 2021 he wrote that he would ‘call s2f invalidated if we have not reached 100K by Dec this year,’ a statement preserved by Protos. December 2021 arrived with Bitcoin under $50,000. The deadline passed. The model was not retired.

That refusal to let the model fail is the thread running through this whole story. Stock-to-flow looked prophetic from 2019 to the middle of 2021, then missed catastrophically the first time it faced a clean test, and its champion waved the miss away. A year on from the October 2025 peak, with Bitcoin sitting at a level the model treats as roughly an order of magnitude too low, the quantitative case against it has hardened from contrarian take into near-consensus. Those arguments are worth laying out one by one, because they double as a field guide for every scarcity story that will follow this one.

What Stock-to-Flow Actually Measures

Stock-to-flow is an old commodities concept, not a crypto invention. ‘Stock’ is the amount of a good that already exists. ‘Flow’ is the amount produced in a year. Divide the first by the second and you get the number of years of current production needed to recreate the entire existing supply. A high ratio means new supply is tiny next to the hoard, the quantitative version of what gold bugs call hardness. Gold scores around 62 on this measure and silver around 22, figures summarized in CoinGecko’s explainer.

Bitcoin’s issuance is fixed in software. Every 210,000 blocks, roughly every four years, the block subsidy paid to miners is cut in half. Because freshly mined coins are the flow, each halving roughly doubles the stock-to-flow ratio in a single block. That mechanical doubling gives the Bitcoin chart its staircase shape and puts the halving at the center of a decade of cyclical lore. If you want the engineering view of where those coins come from, our explainer on what Bitcoin’s miners actually compute walks through the issuance machinery block by block. The table below shows how the ratio climbs across the halving epochs; CoinGecko puts Bitcoin near 27 before the 2020 halving and around 119 after the 2024 halving, with the other values following from the simple fact that flow keeps halving while stock keeps growing.

How PlanB Turned Scarcity Into a Price Line

In March 2019, PlanB published a Medium essay called ‘Modeling Bitcoin’s Value with Scarcity.’ The move that made it famous was deceptively simple. He plotted Bitcoin’s market capitalization against its stock-to-flow ratio on a log-log chart, ran a linear regression, and reported an eye-watering fit of around 95% R-squared. On that chart the data points marched up a near-straight line, and the line implied that each halving should multiply Bitcoin’s fair value. The first published target, for the period after the 2020 halving, was roughly $55,000, as covered at the time by International Business Times.

Presentation did the rest. PlanB, who has never confirmed his identity but describes himself as a Dutch institutional investor, released a live chart that colored each monthly data point by how long remained until the next halving. The result looked less like a speculative regression and more like a law of physics, a rainbow ribbon climbing from bottom left to top right with Bitcoin’s price dutifully tracking it. For a market starved of valuation anchors, a single tidy equation that promised number would go up, with the math to prove it, was irresistible.

Within a year the chart had spawned imitator dashboards, countdown clocks, and a cottage industry of analysts citing it as settled fact. It also became tribal identity, printed on merchandise and waved around as proof that the only open question was how high Bitcoin would go. That cultural grip matters, because it is part of why the model outlived its own failures.

The Predictions, Measured Against the Tape

In April 2020 PlanB raised the stakes with a second version, the Stock-to-Flow Cross Asset model, or S2FX. It dropped time as a variable, folded gold and silver into the dataset, and sorted Bitcoin’s history into four clusters meant to represent phase transitions, from proof of concept to payment network to e-gold to financial asset. The headline output was a market value near $5.5 trillion for the post-2020 period, which worked out to roughly $288,000 per coin before 2024, the target CryptoSlate reported as the model’s defining call. Extended along the same slope, the framework pointed at a seven-figure Bitcoin, a trajectory The Block later summarized as the model expecting a coin well on its way to $1 million by 2026.

The costliest call was short-term. During the 2021 bull run PlanB published a floor model, a set of monthly worst-case prices, that called for a floor of $98,000 in November and $135,000 in December 2021. Instead Bitcoin peaked just under $69,000 in November, closed that month near $57,000, and finished December around $47,000, as documented by U.Today. These were not central estimates that happened to slip. They were the floors, the levels the model insisted Bitcoin would not break below, circulated as reassurance for buyers near the top. The table below lines up the headline forecasts against what the market actually delivered.

The Day Scarcity Didn’t See Coming

If the top exposed what stock-to-flow got wrong about the trend, the crash four days later exposed what it cannot model at all. On 10 October 2025, a post from President Donald Trump threatening a 100% tariff on Chinese imports hit a market stuffed with leverage, and the result was the largest deleveraging in crypto history. CoinGecko’s account of the day describes more than $19 billion in positions liquidated inside 24 hours, roughly nine times any previous single-day total, with Bitcoin falling about 16% from its highs as market makers pulled their bids and thin order books amplified every sell order.

Nothing in a scarcity model anticipates a day like that. The stock-to-flow ratio did not move; the supply schedule was exactly where the code said it would be. What moved was leverage, policy, and liquidity, the forces that actually set prices minute to minute. The crash was a reminder that most of Bitcoin’s short-term volatility lives in the derivatives stack, a structure we unpack in our guide to how perpetual-futures DEXs work. A model that tracks only the issuance curve is blind to the exact machinery that produced the most consequential day of the cycle.

The Shrinking Cycle

Stock-to-flow’s deepest promise is that each halving should matter more than the last, because the ratio jumps by a larger absolute amount every cycle. The price record says the opposite. Measured peak to peak, Bitcoin’s cycle tops have grown by shrinking multiples. From roughly $1,150 in 2013 to about $19,800 in 2017 was a gain of around 17 times. From there to roughly $69,000 in 2021 was about 3.5 times. From there to the $126,198 high of October 2025 was about 1.8 times, the weakest cycle multiple on record, using the peak figures tracked by Coinlaw.

That is the exact inverse of what a scarcity-driven parabola predicts. As Bitcoin’s market value grows, each new wave of buying moves the price less, a straightforward consequence of a larger base, and no tightening of the supply schedule has reversed it. The mining industry has absorbed the same lesson. A year after the top, the companies that bet hardest on endless appreciation are grinding through thin margins, a story we told in our comparison of Marathon and Riot one year after the peak. The table below shows the diminishing returns cycle by cycle.

Spurious Correlation, the Original Sin

The deepest problem with stock-to-flow is not that it missed in 2021. It is that the 95% fit was close to meaningless from the first day. Both Bitcoin’s price and its stock-to-flow ratio rise over time. Regress any two series that both trend upward and you will get a high R-squared whether or not one causes the other; this is the textbook trap of spurious correlation, and economists have warned about it for a century. As early as June 2020, CoinDesk spelled out why fitting price to a steadily rising supply schedule proves very little.

The most cited technical takedown came from the analyst Nick Emblow, whose work under the BTConometrics banner was catalogued by trader Eric Wall in a widely shared post listing the greatest blows to the model. Emblow found the correlation was ‘entirely spurious,’ showing that ‘after accounting for the autocorrelation in the residuals the coefficient for log s2f becomes almost zero,’ which is a precise way of saying scarcity stops explaining anything once the statistics are done properly. A second objection is even more basic: because market capitalization equals price times supply, and supply is the dominant term in stock-to-flow, the regression is partly plotting stock against itself.

A quick thought experiment shows the trap. Regress Bitcoin’s price against almost any quantity that only ever rises, cumulative days since launch, total blocks mined, even the growth of global data storage, and you will land on a similarly flattering line. The fit measures shared upward drift, not a causal link from scarcity to value. The burden was always on the model to prove the relationship was more than two arrows pointing the same way, and that is precisely the burden the next defense was meant to carry.

The Cointegration Fight

Defenders of the model had a comeback. In econometrics, two trending series can be linked legitimately if they are cointegrated, meaning they share a long-run equilibrium that stops them drifting apart forever. PlanB and allied researchers pointed to cointegration tests as proof that stock-to-flow escaped the spurious-correlation trap. For a while that sounded like a decisive rebuttal, and it was repeated often enough to settle the argument in many readers’ minds.

It did not survive scrutiny. In a detailed review for Amdax Asset Management titled ‘The Fall of Cointegration,’ analyst Marcel Burger showed that the original cointegration tests had been applied incorrectly. Sebastian Kripfganz, an econometrics specialist at the University of Exeter and co-author of the widely used ARDL estimation routines, concluded that the specific test invoked was not valid for this setup. Once the deterministic trend baked into Bitcoin’s supply schedule is handled correctly, no statistically sound long-run relationship between stock-to-flow and price survives.

The pattern repeats across every statistical defense of the model. Each one held just long enough to reassure believers, then fell when specialists checked the math. By the time the cointegration claim collapsed, the burden had quietly shifted: defenders were no longer showing that scarcity drives price, only that the opposite had not been proven to their satisfaction. That is a long way from the confident law of nature the original chart seemed to promise.

The Chameleon Problem and Gold’s Tell

The most quotable critique belongs to Nico Cordeiro, chief investment officer at the quant fund Strix Leviathan, whose essay ‘A Chameleon Model’ borrowed a term from Stanford economist Paul Pfleiderer for theories that look rigorous while resting on assumptions that collapse under examination. Cordeiro’s sharpest evidence was gold itself. Over about 115 years, gold’s stock-to-flow ratio held roughly constant near 60, yet its market value swung from around $60 billion to roughly $9 trillion. If a stable ratio is compatible with a hundredfold change in value, the ratio cannot be the thing setting the price.

Cordeiro traced roughly 88% of gold’s long-run value changes to the declining purchasing power of the dollar rather than to any supply ratio, and he dismissed the Bitcoin paper as math-laden marketing. He was blunt about its forecasting value, telling Cointelegraph that the model’s accuracy would likely be ‘about as successful at forecasting Bitcoin’s future price as the astrological models of the past were at predicting financial outcomes.’ The point is not that scarcity is irrelevant to Bitcoin. It is that a ratio which demonstrably fails to price gold, the original hard asset, has no automatic claim to price its digital descendant.

Scarcity Is Not Demand

Every version of stock-to-flow shares one structural blind spot: it models only supply. The price of anything is set where supply meets demand, and the halving changes only the first half of that equation. A fixed, perfectly predictable drop in new issuance cannot by itself dictate price unless you also assume demand keeps rising to absorb it, an assumption the model never states and never tests. Scarcity is a property of the supply curve; it says nothing about how many people want the asset, or at what price.

The last cycle made the point vividly. The rally that carried Bitcoin to its 2024 and 2025 highs owed far more to a demand shock than to a supply one: the arrival of United States spot Bitcoin exchange-traded funds. The Securities and Exchange Commission approved the first eleven spot Bitcoin ETFs on 10 January 2024, a 3-2 decision that Chair Gary Gensler called ‘the most sustainable path forward’ even as he stressed it was no endorsement of crypto. Those funds absorbed tens of billions of dollars of new buying that no scarcity schedule anticipated. Demand, not the halving clock, wrote the recent record, a theme we traced in our look at how macro and political events, not slogans, actually move crypto.

Macro conditions did the rest. Bitcoin’s biggest moves of the last two years lined up with shifts in interest-rate expectations, dollar liquidity, and risk appetite across equities, none of which appear anywhere on a stock-to-flow chart. When policy tightened or a tariff headline landed, Bitcoin fell with other risk assets regardless of where the halving clock pointed. A supply-only model is blind to exactly the forces that dominate real price action.

When a Model Cannot Be Wrong

The most consequential critique is philosophical. A forecast earns trust by being falsifiable, which means it has to be capable of failing. Stock-to-flow kept shedding its failure conditions. When the floor model broke, it was reframed as something separate from the real stock-to-flow model. When the $100,000 deadline lapsed, the model lived on. Each miss was absorbed rather than scored, and every delay became evidence that the big move was still ahead.

Ethereum co-founder Vitalik Buterin made the point the week Bitcoin slid toward $20,000 in June 2022. Stock-to-flow, he wrote, was ‘really not looking good now,’ and he went further: ‘I think 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 reported by Decrypt. He also argued that the broad claim that halvings cause price rises is unfalsifiable, because believers can always point to some later date when the prophecy will finally come true. A model you refuse to abandon after it misses is not a forecast. It is an article of faith wearing a lab coat.

The Power Law, a Sturdier Straw?

As stock-to-flow lost credibility, many of its former adherents migrated to a different single-line model: the Bitcoin power law. Credited largely to physicist Giovanni Santostasi, with a corridor version elaborated by Harold Christopher Burger, it fits price not to scarcity but to time, arguing that Bitcoin’s price has grown as a steep power of its age since genesis and traces a straight line on a log-log chart. Decrypt’s primer lays out the idea, which Santostasi has extended to argue for a multi-million-dollar coin over the coming two decades.

The power law is harder to wave away than stock-to-flow, because it does not predict infinity and because, on past data, it has drawn Bitcoin’s long-run trend with uncanny neatness. But it inherits the same original sin. Time, like scarcity, is not a mechanism. Fitting price to the calendar still regresses one trending series on another, and the model offers no account of the buyers and sellers who actually set prices. Network scientists have also shown that genuine power-law behavior is rarer than enthusiasts assume; a well-known survey led by Aaron Clauset found that only a small minority of real-world networks pass strict power-law tests.

The deeper lesson is that stock-to-flow and the power law are cousins, not opposites. One fits price to scarcity and the other to time, but both reduce a messy, demand-driven market to a single deterministic line and promise a destination. Critics who dismantled stock-to-flow have made the same objection to the power law: plotting price against a variable that only ever rises is not the same as explaining it. Any model that can draw your future from one input is selling confidence, not insight.

The $500,000 Question

PlanB has not recanted. His current stance, reported by Yahoo Finance, is that stock-to-flow still points to an average near $500,000 across the 2024 to 2028 cycle. Set that against the tape and the scale of the claim comes into focus. At almost 20 million coins outstanding, $500,000 each implies a market value around $10 trillion, roughly six times where Bitcoin trades today and larger than the entire above-ground gold market. The model is not asking for a good year; it is asking for a repricing with few precedents in financial history.

The extrapolation gets stranger at the edges. As each halving pushes issuance toward zero, the stock-to-flow ratio races toward infinity, and so does the model’s implied price. A framework that mechanically outputs an infinite valuation is not making an economic claim; it is exposing a math artifact. Gold’s annual production has been small relative to its stock for centuries without its price running away, because scarcity was never the whole story. The deeper you look at where the curve points, the clearer it becomes that stock-to-flow mistook an accounting fact about issuance for a theory of worth, and the cost of mining those coins, detailed in our breakdown of what it really costs to own a miner, sets a far more grounded floor than any scarcity ratio.

How to Read Any Bitcoin Price Model

Stock-to-flow is worth studying precisely because its failures are so instructive. The same red flags recur in nearly every model that promises a clean line to riches.

• Be suspicious of single-variable models. Price is a tug of war between supply and demand, and any framework that tracks only one side is describing half a market.
• Distrust very high R-squared on trending data. When two things both rise over time, a tight fit is the default, not evidence of cause.
• Ask what would prove the model wrong. If every outcome can be explained after the fact, the model is unfalsifiable and tells you nothing in advance.
• Check the edges. A model that implies an infinite price, or any impossible value at its limits, is built on an identity rather than a mechanism.
• Treat a curve as a description, not a destiny. Fits summarize the past; they do not bind the future, and they say nothing about the ETF flows, rate decisions, and liquidity swings that actually move the tape.

None of this means Bitcoin’s long-run case rests on a debunked chart. The credible investment thesis has always been about adoption, monetary policy, and demand, not about a supply ratio marching to infinity. Stock-to-flow’s real legacy may be as a cautionary tale, proof that a pretty equation and a 95% fit can hypnotize an entire market, right up to the moment the deadline it set for itself came and went.

Frequently Asked Questions

What is the Bitcoin stock-to-flow model?

The stock-to-flow (S2F) model, popularized by the pseudonymous analyst PlanB in 2019, estimates Bitcoin’s value from its scarcity. It divides the existing supply, the stock, by annual new issuance, the flow, and uses a regression to map that ratio to price. Because each halving cuts issuance, the ratio roughly doubles every four years, which the model reads as rising fair value.

Why did the stock-to-flow model fail?

It failed on two levels. Statistically, it fit price to a steadily rising supply schedule, a setup that produces a high correlation whether or not scarcity drives value, and proper econometric tests found no sound long-run relationship. Practically, its 2021 floor predictions of $98,000 and $135,000 missed badly, and its path toward a million-dollar coin by 2026 sits roughly ten times above Bitcoin’s actual October 2026 price near $82,000.

What did PlanB predict and did it come true?

PlanB’s original 2019 model implied about $55,000, his 2020 cross-asset version targeted $288,000 before 2024, and the extended trajectory pointed toward $1 million by around 2026. Bitcoin did trade into the $55,000 range in 2021 but never reached $288,000, and in 2026 it sits near $82,000. His 2021 monthly floor model, which called for $98,000 in November and $135,000 in December 2021, missed by wide margins.

Is the Bitcoin power law model better than stock-to-flow?

The power law, associated with physicist Giovanni Santostasi, is better behaved because it does not imply an infinite price and has tracked Bitcoin’s long-run trend closely. But it shares stock-to-flow’s core weakness: it fits price to time rather than to supply and demand, so it describes a trend without explaining what causes it. It is a tidier curve, not a verified theory of value.

What actually drives Bitcoin’s price if not scarcity?

Demand and liquidity do most of the work. The 2024 and 2025 rally tracked tens of billions of dollars of inflows into United States spot Bitcoin ETFs approved by the SEC in January 2024, alongside interest-rate expectations and broad risk appetite. Scarcity sets the supply schedule, but marginal buyers and sellers, not the halving calendar, set the price.

By the HOGE Wire markets desk.