69%.
That was the figure that opened the analysis. Bitcoin's supply in profit โ the share of all circulating coins whose last on-chain movement priced them below the current market โ had risen to 69%. The analyst, publishing under the handle AxelAdler Jr, anchored the claim with a second data point: a 41% change across a 90-day window. Together, the two numbers were presented as confirmation of a short-squeeze-driven repricing, and as a warning that the buying momentum was about to exhaust itself.
Every anomaly is a story the data forgot to tell. So I ran the arithmetic before I ran the thesis.
69 minus 41 equals 28. If "41%" meant a relative gain โ a 41% increase off the prior base โ then 28% ร 1.41 lands at 39.5%, not 69%. The figures do not reconcile under that reading. The only interpretation that survives the arithmetic is that 41 is measured in percentage points. Which means that ninety days before the article, roughly 28% of Bitcoin's supply sat in profit.
That is not a mild correction. That is a capitulation print. And there, inside the first paragraph of a statistic that sounded like a celebration, the story quietly broke.
What Supply in Profit Actually Measures
Before any conclusion can be audited, the instrument has to be understood on its own terms. Supply in Profit, or SIP, is a UTXO-based metric. The calculation walks every unspent transaction output, assigns to each one a cost basis equal to the price at which it last moved on-chain, and compares that basis to the current spot price. The result โ the proportion of supply whose basis sits below spot โ is what the analyst reported as 69%.
The metric is mature and widely used. It shares its lineage with MVRV, NUPL, and Realized Price, and both Glassnode and CryptoQuant publish standardized versions. A determined reader can reconstruct most of it independently from raw chain data. On the surface, that transparency is a virtue.
But three properties of the instrument matter far more than its popularity, and the article addressed none of them.
First, SIP is synchronous at best and lagging in practice. The price moves, and the metric follows. It is a rearview mirror. The instruments that actually precede price โ funding rates, open interest, liquidation cascades โ live on the derivatives side, and SIP contains none of them.
Second, its causality runs in one direction only. Price determines the metric. The metric does not determine price. An analyst who presents SIP as a forecast is reversing the arrow of causation, whether deliberately or through carelessness, and that reversal is the most common analytical error in the entire on-chain discipline.
Third, the adjusted and raw variants of SIP differ by as much as 5 to 15 percentage points, depending on how lost and long-dormant coins are treated. The article disclosed neither its provider nor its variant. Without that provenance, "69%" is a number without a birth certificate.
The ledger doesn't trade on reputation. It trades on reconciliation. And this reconciliation โ at the level of stated methodology โ was never performed.
The Arithmetic That Contradicts the Framing
Internal consistency is the fastest lie detector available to anyone working with numbers, and it costs thirty seconds. Three stated facts frame the puzzle: profit supply at 69%, a 90-day change of 41 percentage points, and a spot price of approximately $77,000.
If 90 days ago 28% of supply was in profit, and today the price is $77,000 with 69% in profit, then the remaining 31% of supply carries a cost basis above $77,000. That residue is more interesting than the headline. Thirty-one percent of the coin supply bought higher than the current price.
Those holders are, almost by definition, short-term holders โ the cohort that acquired during the previous run-up and now sits underwater. Their behavior is the single most important variable for what happens next, and the article never segmented them out. It offered a total and called it a signal.
This is the difference between a sum and a distribution. SIP compresses the entire holder base into one number and discards the question that actually matters: whose coins are in loss, and how quickly do those specific holders sell when the market hands them their money back?
I learned this lesson from the other side of the trade. When I built my Python backtesting engine through the 2020 DeFi Summer, I simulated yield-farming strategies across Compound and Uniswap and processed more than 10,000 swap events. Aggregate liquidity looked bottomless on every dashboard. Slippage told a different story: the apparent arbitrage in early Aave deployments was routinely erased by MEV bots before a retail hand could reach it. The distribution underneath the aggregate is always where the risk hides. Every dashboard shows the sum. The loss is always in the tail.
The Date Problem
There is a second anomaly, quieter than the arithmetic but more corrosive. The article is dated "September 11" with no year attached.
That omission is not cosmetic. Bitcoin's price history is a sequence of regimes, and a $77,000 spot price combined with a capitulation-level 28% profit supply ninety days prior does not map cleanly onto any single well-known calendar window. If "September 11" refers to a year in which Bitcoin traded at $77,000, the implied 90-day path from a 28% profit-supply trough is extraordinary. If it refers to a different year, the price figures and the profit-supply figures describe two different worlds stitched into one narrative.
An analyst who publishes a market call without a timestamp is publishing a claim that time itself cannot falsify. The half-life of a tactical market view is measured in days. Strip the year from the dateline and you strip the reader of the only tool that can tell them whether the call is live or already dead.
The Short-Squeeze Mechanism โ Valid in Logic, Unverified in Fact
The analyst's central mechanism is the short squeeze, and the logic chain is sound. It runs like this. Shorts become crowded โ funding turns negative, open interest climbs. Price rises and forces liquidations, and every forced liquidation executes as a market buy. Those buys push price higher and trigger the next tier of liquidations. A positive feedback loop. When the short-side position pool empties, the marginal buyer vanishes. Momentum hits a vacuum, and price retraces part of the move.
This is not a hypothetical shape. January 2023, August 2024, April 2025 โ the same pattern recurs across cycles. As a mechanism, it is well established, and the analyst is right that it exists.
The problem is not the mechanism. The problem is the evidence.
To claim that a squeeze drove a move, an analyst must demonstrate derivatives-side proof: funding rates flipping negative, open interest collapsing, liquidation volume spiking, the long/short account ratio stretched to an extreme. The article provided none of these. Not one. It asserted "mechanical repricing" and moved forward, and that single phrase was asked to carry the weight of a dataset. It cannot.
A claim that cannot be falsified by the data it cites is not analysis. It is narrative wearing the costume of analysis.
Correlation is the ghost; causation is the corpse. The analyst showed us a ghost โ momentum fading โ and declared it the cause of a death that had not yet occurred.
The Supply Wall the Article Touched But Never Explained
Here is where the source material contained its most defensible technical insight and then failed to press it.
A rise in SIP from 28% to 69% is not merely a bullish headline about "profit." It is the construction of a supply wall. As price climbs, holders whose cost basis sits just below the current market โ the newly-profitable cohort โ approach the psychological breakeven line. Many of them waited years to get out even. When the market finally offers them their money back, a meaningful fraction takes it.
The 31% of supply still underwater is the mirror image. Every recovery rally lifts part of that cohort back toward breakeven, and part of it sells into the strength. That is the mechanical source of overhead resistance โ a far stronger and more falsifiable argument than the squeeze thesis the analyst actually leaned on.
This is why liquidity is the oxygen; volatility is the breath. The market can hold a bid indefinitely if the oxygen of real demand keeps flowing. But the breath of a squeeze is always short. The analyst described the breath and mistook it for the oxygen.
Cross-Checking With Sibling Metrics
A rigorous analyst never trusts a single instrument when the same question can be triangulated. SIP has siblings that answer adjacent questions: MVRV, which compares market value to realized value across the whole supply; NUPL, which expresses the net unrealized profit or loss as a share of market cap; and Realized Price, the aggregate cost basis of the network. A proper case โ bullish or bearish โ would present SIP alongside at least one sibling and confirm that they agree.
The article presented SIP alone. If MVRV and NUPL, freshly computed, told a similar story, they would strengthen the case for a maturing rally. If they diverged โ if MVRV suggested the network's aggregate profit was thinner than SIP implied โ the headline 69% would be an outlier rather than a consensus. The reader is never handed the chance to check, because the sibling metrics are never mentioned.
This is not pedantry. It is triangulation, and it is the difference between a single noisy reading and a confirmed signal.
Historical anchoring reinforces the point. Readings in the high twenties on SIP are rare. They cluster in the deep capitulation windows โ the fourth quarter of 2018, the March 2020 liquidity crash, the November 2022 aftermath of FTX. Each of those was a genuine bottom, and each produced a similar-looking trough. If the article's 28% starting point is accurate, the analyst is describing one of the sharpest recoveries off a capitulation floor in Bitcoin's history โ a recovery that deserves far more cautious framing than a "momentum is fading" warning implies. A market climbing that violently out of deep loss has not run its course. It has just woken up.
The Seven Missing Inputs
The absence of derivatives data is not an isolated gap. It is one of seven structural absences that, taken together, render the article's central claim unverifiable. Each of the following was omitted: stablecoin net inflows, the proxy for fresh capital entering the ecosystem; exchange net flows, the immediate reservoir of sell pressure; futures open interest, a necessary condition for identifying a squeeze; funding rate, the second necessary condition; the Fear and Greed index, a quantitative sentiment anchor; ETF net creations and redemptions, the dominant marginal demand source since 2024; and liquidation data, the direct evidence of squeeze intensity.
A conclusion about short-side crowding requires measures of short-side crowding. When the article contains none, the conclusion is capped at the level of an observation. Reasonable, perhaps. Demonstrated, no.
I watched a version of this mistake play out in the worst possible way. Through 2022, my framework monitored TerraUSD's reserve ratio against its on-chain stablecoin supply, and the divergence appeared weeks before the collapse. The lesson was not that data works. The lesson was that the right data works, and that a metric pulled from the wrong layer is a false witness dressed as an expert.
Momentum Fade Is Not Price Correction
The discipline of the economist demands a distinction the article blurred: the difference between a rate and a level.
"Momentum fading" is a statement about the rate of change. It is a high-confidence inference. When a squeeze exhausts the short-side pool, the velocity of buying genuinely decelerates. No serious reader should dispute that.
"Price correcting" is a statement about the level. It requires additional conditions to be met โ macro liquidity tightening, ETF net outflows, miner selling. The article extrapolated the first claim directly into the second, and that leap is where the logic fails.
A rate of change can slow while the level still grinds higher. A squeeze can end and price can simply flatline for weeks. The analyst treated "momentum fading" as a synonym for "price falling." They are not the same sentence. One is arithmetic; the other is a forecast.
The Tool Does Not Match the Target
There is a more fundamental problem, and it is structural rather than arithmetic.
Bitcoin sits at the base of the crypto stack โ settlement layer, reserve asset, the collateral beneath everything else. Its price is disproportionately determined by forces outside the chain: macro liquidity, real interest rates, risk appetite, the path of the Federal Reserve, and the flow of ETF capital. On-chain supply structure is an endogenous variable. The analyst used an endogenous instrument to predict an exogenously driven asset.
This is a mismatch of categories, not a matter of degree. Since the launch of spot ETFs, a large share of Bitcoin has moved into custodial storage and stopped moving entirely. SIP is structurally desensitized to that dormant supply; it cannot see the coins that never move. The metric's marginal information content has been diluted by the very instrument โ the ETF โ that reshaped the market. The analyst did not adjust for it, and may not have known to.
Compounding errors are just debt in disguise. An analysis built on a lagging metric, applied to an exogenously driven asset, with no macro context appended, accumulates interpretive debt with every paragraph. That debt comes due the moment the thesis meets reality.
The Omitted Supply-Side Variable
One more absence deserves naming: miner economics.
After the April 2024 halving, the block subsidy fell to 3.125 BTC. Annualized issuance now sits near 0.85%, and it will fall again toward 0.4% after the next halving. This means the supply side is rigid and predictable โ it cannot manufacture sustained sell pressure on its own. Therefore the "fading momentum" the analyst described is, at its root, a demand-side problem, not a supply-side one.
But the halving also cut miner revenue, and marginal miners facing a higher cost per coin may sell into weakness to cover operating costs. That is a supply-side amplifier that SIP cannot see. Ignore it, and you miss half of the correction mechanism you are trying to describe.
Trust is a variable, not a constant. And so is miner behavior at the margin. Both must be modeled, not assumed.
The Bot in the Order Book
The 2026 dimension compounds every problem above. I have spent the past year modeling the economic behavior of autonomous blockchain agents โ a game-theoretic framework predicting how AI-driven bots interact with decentralized oracle networks under varying reward structures. One conclusion is directly relevant here: as autonomous agents dominate order flow, the fraction of price movement driven by mechanical execution โ liquidation engines, arbitrage bots, funding-rate harvesters โ rises, and the fraction driven by human conviction falls.
A squeeze is precisely a mechanical event. Its predictability is a function of how much of the book is machine-controlled, and its aftermath is a vacuum that humans are slower to fill. The analyst's instinct that momentum fades may actually be sounder in a bot-dominated market than in any prior cycle โ but for reasons the article never articulated, and with timing the article cannot supply.
Where the Correction Actually Lands
Even granting the analyst's conclusion โ momentum fades, price retraces โ the asymmetry of transmission deserves attention the article never gave it.
Bitcoin's beta to the broader market is roughly 1. Altcoins routinely run 1.5 to 2.5. If Bitcoin pulls back 8%, the altcoin complex often prints 15% or worse, and DeFi collateral โ posted largely in these volatile assets โ faces liquidation cascades that compound the drawdown further. The pain radiates outward from the base layer, and it grows louder as it travels.
The trade the analyst implied, if any, is "reduce exposure." That trade has its highest expected value not in Bitcoin, but in the assets that amplify Bitcoin's moves. The analyst stopped at the symptom and never reached the instrument. A market call that identifies the direction but not the vehicle delivers half its potential value and keeps the other half on the table.
Takeaway
The number was 69%. The story attached to it was about a squeeze fading and a price about to correct.
The arithmetic says something narrower and more useful. Thirty-one percent of Bitcoin's supply is underwater at $77,000, and that cohort is the source of the overhead wall. Watch the funding rate and open interest โ the two inputs the article omitted, and the two that would confirm or kill a squeeze thesis outright. Watch ETF flows โ the exogenous demand that decides whether the supply wall absorbs a pullback or collapses under it.
If funding normalizes and open interest falls without a price break, the squeeze ended and the market simply digested the move. If funding stays negative and ETF flows turn outward, the retracement is real.
The ledger will record which one it was. It always does. The only open question is whether anyone reads it before the position closes โ or after.
