PlanB Still Says $500K but Steers by Difficulty, Not S2F

The Clearest Verdict on S2F Comes From Its Own Creator

The sharpest verdict on Bitcoin’s most famous price model did not come from an academic, a short-seller, or a rival analyst. It came from the person who built it. In September 2026, PlanB, the pseudonymous analyst who published the stock-to-flow (S2F) model in 2019 and turned it into the defining chart of the last bull market, told followers that mining difficulty is “a more reliable indicator for Bitcoin price prediction than S2F or power law models,” according to remarks reported by KuCoin. He was not joking. He was describing how he now reads the market: by moving averages, realized price, the relative strength index, and the economics of mining, rather than by the scarcity ratio that made his name.

Sit with that for a moment. The one person with the most to gain from S2F being right has, in practice, stopped steering by it. He still keeps the headline number on the wall, a cycle-average “value signal” near $500,000 that he has reiterated for years, per Yahoo Finance. But the toolkit he actually uses is the same price-versus-time technical kit that S2F’s critics always said was doing the real work.

The timing makes the point hard to miss. On 16 October 2026, Bitcoin’s mining difficulty rose 5.92% to a record 140.58 trillion, the third-largest jump of the year, as network hashrate ran above 1,031 EH/s, per TokenPost. By PlanB’s own framework, that is the constructive signal: rising difficulty, he says, “may indicate miners are increasing their investment and the market bottom is becoming more confirmed.” Notice what carries the weight in that read. Not the halving. Not the 21-million cap. Miner behavior, hashrate, and cost. The author of the scarcity model is watching the furnace, not the vault.

Bitcoin trades near $83,300 as this is written, up about 2% on the day but roughly 34% below its October 2025 record of about $126,000, per CoinGecko. The S2F “value” for this cycle, close to $500,000, implies a market capitalization near $10 trillion, around six times today’s figure. That gap is not a rounding error. It is the story. And the most honest guide to what it means is the behavior of the man who drew the line.

What Stock-to-Flow Actually Claims

Stock-to-flow started as a way to price commodities by scarcity. Stock is the existing above-ground supply; flow is the new amount produced each year. Divide the first by the second and you get how many years of current production it would take to recreate the entire stock. Gold scores high, around 62, because its existing hoard dwarfs annual mine output; silver is lower, near 22, per CoinGecko’s explainer. The higher the ratio, the “harder” the asset is said to be.

PlanB’s 2019 paper, titled “Modeling Bitcoin’s Value with Scarcity,” took that ratio and did something commodity analysts had not: it regressed Bitcoin’s dollar market value directly against its S2F ratio and reported an R-squared around 95%, in the original paper. Because Bitcoin’s supply is fixed in code, each halving roughly doubles the ratio by cutting new issuance in half. After the April 2024 halving, Bitcoin’s S2F sits near 119, already well above gold. Run that through the fitted curve and it returns six-figure prices; a 2020 cross-asset revision, S2FX, put the 2020-2024 average near $288,000 with a reported R-squared of 99.7%, and later work pushed the 2024-2028 figure toward $500,000.

The appeal is obvious. Supply is the one thing about Bitcoin you can know with near-certainty. Issuance is enforced by consensus rules and runs on autopilot toward the final satoshi around the year 2140. If price is a tidy function of a known, shrinking flow, then the future looks knowable. That is the promise. The rest of this piece is about why the promise keeps breaking, and why even its author no longer trades on it.

A $500K Model and an $83K Market

Put the model’s output next to the tape and the distance is stark. After the 2024 halving the block reward fell to 3.125 BTC, roughly 450 new coins a day, an annual supply inflation under 1%. S2F’s logic says that ever-tightening flow should pull price higher with each cycle. Instead, the 2024-2028 cycle has so far produced a record near $126,000, about twice the prior peak, followed by a slide into the low $80,000s.

PlanB’s standing defense, repeated for years, is that the $500,000 figure is a four-year average and a “value signal,” not a timing tool, and that outside shocks rather than model error explain the shortfall, a framing laid out in his 2026 scenarios on Bitbo. Fair enough as far as it goes. But the cycle is most of the way through, the scheduled supply shock arrived exactly on time, and price went the other way.

The $500,000 “value” implies roughly $10 trillion of market capitalization. That is about six times Bitcoin’s current $1.68 trillion and on the order of half the value of all the gold ever mined. None of it is impossible on a long horizon. The problem is that S2F claims to tell you when, keyed to the halving clock, and the clock just chimed with no parade. For a model sold on precision, “eventually” is a heavy retreat.

PlanB’s Real 2026 Toolkit: Moving Averages, Not Scarcity

Here is the part that rarely makes the headlines. Watch what PlanB actually does, day to day, and you will not find much scarcity math. You will find technical analysis.

In February 2026, with Bitcoin sliding, he laid out four bear scenarios. The worst case was an 80% fall from the $126,000 peak to around $25,000. The “typical” bottom he pegged at $50,000 to $60,000, explicitly near the 200-week moving average and realized price, which then stood at about $58,000 and $55,000. A shallower retrace would hold above the prior cycle high near $69,000 to $70,000. A fourth, “already bottomed” scenario near $72,900 was invalidated weeks later when Bitcoin traded down to $69,031, per Bitbo. Every anchor there is a line drawn from price history: a moving average, a realized-price level, a prior high. Not one is derived from the stock-to-flow ratio.

In that same note he conceded something the model’s fans tend to skip. His framework expects peaks in the first or second year after a halving, he said, but “it didn’t happen after 2024 halving.” By September he had gone further, telling followers Bitcoin was “in a deleveraging phase,” that they should wait for difficulty to rise again, and that difficulty beats “S2F or power law models” for reading price, per KuCoin.

This is not a gotcha. Using the tools that work is the responsible thing for any market commentator to do. But it quietly concedes the core critique. If the author of the scarcity model reads the tape with moving averages, realized price, RSI, and miner economics, then the scarcity ratio was never the engine. It was the poster on the wall behind the desk.

Mining Difficulty Just Flashed His Own Bottom Signal

The irony sharpened this month. On 16 October 2026 the network’s difficulty adjustment pushed difficulty up 5.92% to 140.58 trillion, the highest reading since March and the third-largest jump of the year, with hashrate running above 1,031 EH/s after crossing the 1 zettahash-per-second line, per TokenPost. Earlier in 2026 the network had printed three consecutive difficulty decreases, the first such run since July 2022, a textbook capitulation tell as stressed miners switched machines off, according to CoinShares’ 2026 mining report. Difficulty turning back up is the mirror image: hashpower returning, miners committing capital again.

The broader picture frames how unusual that turn is. Of 21 difficulty adjustments in 2026, the network has logged 10 increases and 11 decreases, a tug-of-war that maps the year’s price stress far better than any scarcity ratio did. Hashprice, the daily revenue a miner earns per unit of computing power, had sunk toward a five-year low earlier in the year, leaving an estimated 15% to 20% of older machines unprofitable before the October bounce, per CoinShares. When PlanB says to wait for difficulty to rise, he is really saying to wait for that squeeze to clear.

By PlanB’s framework that is the market “bottom becoming more confirmed,” though he was careful to call it “not an immediate buy signal, but rather an observation of miner capacity cycles,” per KuCoin. Whatever you make of the signal itself, look at the frame. The question is no longer what scarcity says the price should be. It is whether the people who turn electricity into coins are willing to spend more to do it. That is a demand-and-cost question, answered by behavior on the ground, not a supply ratio read off a fixed schedule.

It is also a reminder that the hashrate rests on a real, physical industry with its own cycles. The machines, the power contracts, and the mining hardware behind Bitcoin’s energy and e-waste footprint move with price in ways a scarcity chart cannot capture, and they are exactly what PlanB is now watching.

Production Cost, Not Scarcity, Is the Floor That Held

If you want a supply-side number that actually tracked Bitcoin’s 2026 lows, skip the scarcity ratio and look at what it costs to mine a coin. CoinShares put the weighted-average cash cost per Bitcoin mined by public miners near $79,995 in the fourth quarter of 2025, in its 2026 mining report. By early February 2026, with Bitcoin down toward $70,000, CoinDesk reported the average all-in cost to mine a coin around $87,000 per Checkonchain data, meaning price had fallen roughly 20% below production cost.

The network did what mining networks do under stress: hashrate fell about 10% from its October peak, difficulty dropped, weaker machines went dark, and the cost of producing the next coin reset lower until price and cost converged again. That is a real feedback loop with a causal mechanism. When price falls below cost, miners capitulate, difficulty falls, production cost falls, and sell pressure eases; when price recovers, hashpower and difficulty climb back, as they did this month.

Charles Edwards’ energy-value and production-cost work has leaned on exactly this logic for years, and it is why PlanB himself now watches difficulty. S2F has no such mechanism. It does not know or care what a coin costs to produce; it only counts how many exist. When the one supply-side gauge that moved with the 2026 market is about cost and difficulty rather than scarcity, the scarcity story has a problem.

Just Price vs Time: The Whole Critique in Four Words

Strip away the commodity analogy and S2F has a statistical problem that critics flagged almost immediately. Cory Klippsten, chief executive of Swan Bitcoin and a long-time skeptic, put it plainly: the model shows a high correlation because “it uses the same term in each axis,” and once you account for that, it “does not need S2F at all. Just price vs time,” he told Bitcoin Magazine.

The mechanics are simple. A market value is, by definition, supply multiplied by price. S2F regresses that market value against the stock-to-flow ratio, which is itself built from supply. Because supply only ever grinds upward on a fixed path, and price has also trended upward across Bitcoin’s life, you are largely regressing one rising series against another. That produces a gorgeous R-squared and almost no information. Adjust for the autocorrelation and the apparent fit collapses toward zero; independent analyses found no statistical cointegration between the ratio and price.

This is why “just price vs time” is the whole game. If a model’s explanatory power comes from the shared upward drift of two time series rather than from any causal link between scarcity and value, it will look brilliant in-sample and fail the moment the future stops rhyming with the past. After 2021, it did exactly that.

The Chameleon That Keeps Changing Its Target

A good model makes a call and lives with it. S2F keeps getting new clothes. Nico Cordeiro, then chief investment officer at Strix Leviathan, dissected this in a widely cited 2020 note, borrowing Stanford economist Paul Pfleiderer’s term: S2F is a “chameleon” model, one whose assumptions are quietly adjusted until it matches whatever reality has already delivered. His verdict, carried by CoinDesk, was that the model was “about as successful at forecasting bitcoin’s price as the astrological models of the past were at predicting the future.”

The receipts back him up. The original 2019 model implied a 2020-2024 average near $55,000. The 2020 cross-asset revision jumped the target to $288,000 with an even higher reported R-squared, which is less a badge of accuracy than a sign of a curve fitted so tightly it is almost certainly overfitted. The current figure sits near $500,000. Each revision arrived after the prior one strained against the tape, and each pushed the payoff further into the future.

Defenders have a ready answer: the model is not wrong, it is early, and the four-year average will pull back toward the line before the cycle ends. The trouble with being perpetually early is that it is indistinguishable from being wrong. A target that always sits a few years out, and that slides further out each time the deadline nears, cannot be tested by any observation you could make today. That is not a forecast with error bars. It is a promise with an adjustable date, and an adjustable date is the opposite of a prediction.

A model you can always re-fit to the last data point is unfalsifiable, and an unfalsifiable price target is a horoscope with a spreadsheet. It will never be wrong, because “wrong” has been defined out of existence. That is comforting to hold and useless to trade.

The One Thing S2F Got Right

To be fair to the model, give it the credit it is owed. S2F made exactly one promise that held: the supply schedule. The April 2024 halving cut the block reward to 3.125 BTC on the block it was coded to, new issuance fell to roughly 450 coins a day, and annual supply inflation dropped below 1% without a committee meeting or a vote. That is a genuinely remarkable property, and it is the kernel of truth that makes the rest of the model feel plausible. Bitcoin really is the hardest-supply asset ever built, and its issuance really is more predictable than any central bank’s.

But that is also the trap. The halving arrived precisely as advertised, and price did the opposite of what the model said it would. If the one input S2F gets right is the supply shock, and the supply shock fired on schedule while price fell, then the model has run its cleanest possible experiment and failed it. Getting the easy half of the equation right does not excuse being blind to the hard half. A weather model that nails the calendar date of the solstice but cannot tell you whether it will rain is not a weather model. It is a calendar.

The Demand Blind Spot: Six Forces, S2F Sees One

S2F models exactly one variable: supply. It is structurally blind to the other side of every trade. Consider what actually moved Bitcoin across the 2024-2026 cycle.

The halving cut issuance on schedule, and S2F saw it. But US spot Bitcoin ETFs launched and became one of the largest holders of the asset; corporate and sovereign treasuries kept accumulating; the Federal Reserve under Kevin Warsh raised rates in September 2026 for the first time since 2023; Treasury operations reshaped dollar liquidity; and on 10 October 2025 a record wave of liquidations, roughly $19 billion, erased value in a single session, as CoinShares documented. S2F sees exactly one of those forces, the one already written into the code. As the companion case for liquidity beating scarcity argues, the marginal buyer and the cost of money now set the price, not the issuance schedule.

This is why the halving “worked” and the price did not follow. The supply event arrived precisely as coded. The demand that was supposed to meet a shrinking float showed up as ETF flows and macro liquidity, which surge and vanish on their own clock. A model that can see only one of six levers will be right about the lever it sees and silent on the five that matter more.

The Scoreboard: Every Documented Call

Reputations are made on calls, so here is the record. The most concrete S2F-adjacent call was the 2021 “floor model,” which projected rising monthly price floors. It hit for August, September, and October 2021, then called for a $98,000 floor in November and $135,000 in December. Bitcoin closed both months far below, near $47,000, and PlanB conceded on 24 December 2021: “Floor model first miss (after nailing Aug, Sep, Oct). No model is perfect, but this is a big miss and the first in 10y,” as CryptoPotato recorded. The core S2F line fared no better through 2022, when Bitcoin fell under $16,000, more than 80% below the model’s cycle floor.

There is a deeper pattern the model gets backwards. Each Bitcoin cycle has produced a smaller peak-to-peak multiple than the last: tens of times over from 2013 to 2017, about 3.5 times from 2017 to 2021, and only about 1.8 times from 2021 to 2025, as CoinDesk has charted. S2F’s geometry implies the opposite, that each halving should produce a more extreme move as the ratio doubles. The data says the law of large numbers is winning: a $1.7 trillion asset cannot repeat the percentage gains of a $10 billion one, no matter how hard its supply becomes. Whether that means the classic four-year rhythm is over or merely flattening is the subject of its own halving cycle math.

Why a Failed Model Refuses to Die

If S2F is this broken, why is it still everywhere? Part of the answer is demand, not for Bitcoin but for certainty. A clean line that says the number goes up, on schedule, is emotionally and commercially irresistible in a market defined by doubt. Part is unfalsifiability: because the target is a multi-year average that can be re-fit, there is always a reason this year does not count and the line still holds. Part is identity, because people who bought the thesis bought a worldview, and worldviews do not update on one bad quarter.

There is also a reflexive reason the model worked once and may never again. In 2020 and 2021, enough buyers believed the S2F line that their conviction helped push price toward it, a self-fulfilling Schelling point where the chart became a coordination device for a mostly retail market. That is the closest thing S2F has to a success, and it is not a vindication of scarcity so much as proof that a shared story can move price for a while. The marginal buyer in 2026 is an ETF desk or a corporate treasury, not a retail trader studying a rainbow chart, so the one mechanism that ever made the model self-fulfilling has quietly been retired.

And part is that its creator never fully recanted. PlanB has conceded specific misses, including the floor model, and has acknowledged that the four-year cycle “didn’t happen after 2024 halving,” yet he still “reiterated his stock-to-flow value signal remains around $500,000,” per Bitbo. That is the move that keeps the model alive: retire it as a working tool, retain it as a destination. The number survives precisely because it has been redefined into something that cannot be checked until 2028, and probably not cleanly even then.

Matt Hougan, chief investment officer at Bitwise, framed the broader shift bluntly at the end of 2025, arguing the four-year cycle is giving way to a longer, steadier grind as institutions replace retail as the marginal buyer, in a memo covered by CoinDesk. Vitalik Buterin’s warning from 2022 has aged just as well: 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,” he told The Block. The mockery is not the point. The false certainty is.

The Heirs Face the Same Test

S2F’s decline left a vacancy, and several models moved to fill it. The most prominent is the Bitcoin power law, which fits price to time since genesis raised to a high power. It has aged better than S2F, partly because it does not pretend the halving matters and partly because a straight line on a log-log chart is forgiving. But it fails the same core test: it is still price as a function of time, exactly the thing Klippsten said was doing the work in S2F, and its bands are wide enough that almost any outcome can be said to “confirm” it.

A healthier family of tools does not try to name a price at all. On-chain gauges such as MVRV, realized price, and NVT describe where the market sits relative to its aggregate cost basis, cheap or expensive, euphoric or capitulated, without promising a destination. That restraint is the point. A thermometer that tells you it is hot is more useful than an astrologer who tells you the exact temperature next spring.

This is also where PlanB landed. Realized price and moving averages, the anchors in his 2026 bear scenarios, are descriptive, not predictive. They tell you where the market is, not where it must go. The models that survived S2F’s fall are the ones honest enough to give a reading instead of a prophecy, which is a quieter and far more useful thing to own.

How to Read Any Price Model (and What PlanB’s Pivot Teaches)

The practical lesson is not to ignore all models. It is to run any model through a few hard questions before you let it size a position. Does it predict out of sample, or only fit the past? Does it escape the price-versus-time trap, or is its fit just shared drift? Does it model demand, or only supply? Is it falsifiable on a horizon you will actually live through, or does the payoff always recede? S2F fails all four. The power law fails most. Production-cost and difficulty models at least answer a causal question, which is why a scarcity-model author reaches for them when real money is at stake.

Near term, the forces that will set Bitcoin’s price are not on any halving chart. The Federal Reserve meets on 27-28 October 2026, with markets pricing roughly an 80% chance of a hold and a live rate hike penciled in for December; Chair Warsh has pointedly refused to pre-commit, as CNBC reported. ETF flows, dollar liquidity, and the next difficulty adjustment will matter more to the next 20% move than the stock-to-flow ratio will to the next five years.

The surest sign of that is not in any critic’s spreadsheet. It is in the fact that the analyst who drew the most famous scarcity line in crypto now spends his time watching moving averages and the cost of a kilowatt-hour. When the mapmaker stops using the map, it is worth asking what he is navigating by instead.

Frequently Asked Questions

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

It is a scarcity-based price model published by the pseudonymous analyst PlanB in 2019. It divides existing supply (stock) by annual new issuance (flow) and fits Bitcoin’s market value to that ratio, implying higher prices as each halving cuts new supply. Its best-known output is a 2024-2028 cycle value near $500,000.

Has PlanB abandoned stock-to-flow?

Not officially. He still cites a $500,000 S2F value signal as a long-run cycle average. In practice, though, he now reads the market with moving averages, realized price, RSI, and mining difficulty, and said in September 2026 that difficulty is a more reliable indicator for price than S2F or power law models. The method changed even if the headline number did not.

Why has the stock-to-flow model been wrong?

Three reasons. Statistically it largely regresses two upward-trending series, a supply-based ratio and price, so its high R-squared reflects shared drift rather than causation. It has been re-fitted repeatedly, from about $55,000 to $288,000 to $500,000. And it models only supply, ignoring the demand forces, such as ETF flows, macro liquidity, and the Fed, that actually set price.

What is Bitcoin’s stock-to-flow value in 2026?

After the 2024 halving Bitcoin’s S2F ratio is about 119, above gold’s 62. The model’s implied cycle value is near $500,000, which would require roughly $10 trillion in market capitalization, about six times the $83,000 price and $1.68 trillion market cap seen in October 2026.

What works better than stock-to-flow for pricing Bitcoin?

No model reliably predicts price. Production-cost and mining-difficulty measures at least answer a causal question, what it costs to produce a coin, and tracked the 2026 lows better than scarcity did. On-chain gauges like MVRV and realized price describe whether the market is cheap or expensive without naming a target, which is a strength rather than a weakness.

Priya Reddy is a markets writer at HOGE Wire covering Bitcoin, macro, and the models people use to forecast both.