On October 4, Glassnode published a chart that the crypto media ecosystem promptly compressed into a single, shareable sentence: Bitcoin's Long-Term Holder MVRV has not dipped below 1.0 at any point during this cycle — the first time that has happened since 2015. The implication traveled faster than the methodology ever could. Long-term holders have never been underwater. The bottom is behind us. This cycle is structurally different. I have spent the better part of a decade auditing claims like this one, and I can tell you the sentence is technically true and analytically almost empty in isolation. A ratio that has never crossed a line tells you about the denominator as much as it tells you about the numerator — and Glassnode's denominator is a moving target by construction. What follows is a teardown, not a takedown. The signal is real. The story being built on top of it is not yet earned.
The report in question is a single-source on-chain insight from Glassnode, one of the two or three data providers that the institutional side of this market actually treats as infrastructure. That pedigree matters, and I will give it its due. But a data provider's reputation is not a substitute for a disclosed methodology, and a viral chart is not a proof. Audit the code, not the pitch — and in on-chain analytics, the "code" is the cohort definition, the realized-price algorithm, and the address-clustering heuristic. None of the three is fully public. That is the first red flag, and it is not a small one.
Context: What LTH-MVRV Actually Measures, and Why the 155-Day Line Is Load-Bearing
To dissect the claim, you have to understand the machinery that produces it. MVRV stands for Market Value to Realized Value. The numerator is trivial: it is simply the current market capitalization, spot price multiplied by circulating supply. The denominator is where the intellectual work lives. Realized Value — sometimes called Realized Capitalization — is computed by taking every coin and valuing it not at today's price but at the price at which it last moved on-chain. A coin that has sat untouched in a wallet since 2017 is counted at its 2017 acquisition price; a coin that moved this morning is counted at this morning's price. Sum the whole set and you get an approximation of the aggregate cost basis of the network — what the entire market collectively paid for the coins it currently holds.
The metric was formalized by Murad Mahmudov and David Puell in 2018, building on the realized-capitalization concept that had been circulating in the on-chain community. The insight was elegant: if you divide what the market says the coins are worth today by what the market paid for them, you get a single number that expresses whether the average holder is in profit or in loss. MVRV above 1 means the network as a whole is sitting on unrealized gains. MVRV below 1 means the network as a whole is sitting on unrealized losses. It is a sentiment gauge rendered in arithmetic rather than in survey data, and it has an unusually good historical track record of marking cycle extremes — high values clustering near euphoric tops, values below 1 clustering near capitulation bottoms.
LTH-MVRV is a cohort slice of that same computation. Instead of applying the ratio to every coin, you apply it only to the coins held by Long-Term Holders. And this is where the first load-bearing assumption enters, because "Long-Term Holder" is not a natural category. It is a definition. Glassnode's standard definition classifies any address that has held its coins for at least 155 days without spending them as a Long-Term Holder. Cross that threshold and you graduate from Short-Term Holder to Long-Term Holder; spend and you reset the clock. The 155-day line is not derived from first principles. It is a convention, chosen because it roughly corresponds to the historical duration of a bear-market drawdown, and because it produces a clean separation between the cohort that trades the cycle and the cohort that sits through it.
I want to be precise about what this means, because the imprecision is exactly where the headline lives. LTH is not a fixed set of disciplined believers. It is a rolling, dynamic bucket that continuously absorbs whichever short-term holders manage to survive long enough to age into it. The cohort is not selected for conviction; it is selected for endurance. And endurance, in a market that trends down for long stretches, is a filter that behaves in a very particular and very misleading way. Hold that thought. It is the spine of the entire critique.
The Denominator Nobody Reads: Decomposing the Ratio
Here is the analytical move that almost every viral summary of this chart skips. LTH-MVRV is a ratio, and a ratio can stay above a threshold for two completely different reasons. The numerator can rise, or the denominator can rise, or both can move together. When someone tells you "LTH-MVRV never broke 1 this cycle," they are telling you the ratio never broke 1. They are not telling you which component did the work, and the two stories have opposite implications.
Start with the numerator effect. The numerator is spot price. If price simply never fell far enough, for long enough, to push the average long-term holder into loss, then LTH-MVRV stays above 1 and the bullish reading is at least partly honest. This cycle, Bitcoin's drawdowns have been shallower and shorter than in 2015, 2018, or 2022. The price never spent the kind of grinding, multi-month stretch below the aggregate cost basis that characterized previous bear phases. On that reading, the anomaly is a statement about price resilience, and it dovetails neatly with the "institutional floor" narrative that has dominated this cycle's commentary.
Now the denominator effect, which is the one that gets buried. The denominator is Realized Value — the aggregate cost basis of the long-term holder cohort. And here is the mechanical fact that changes the entire interpretation: the long-term holder cohort's cost basis rises automatically during a bear market, because the coins that age into the cohort are precisely the coins that were bought at higher prices and then held through the decline. Think through the plumbing. During a drawdown, short-term holders who capitulate sell their coins. Those coins move to new owners and their realized price resets to the low, capitulation price. But the short-term holders who do not capitulate — who sit frozen through the pain — eventually cross the 155-day threshold and graduate into the LTH bucket, carrying their original, higher cost basis with them. The cohort does not just grow; it grows by importing expensive coins.
This is not a subtle statistical footnote. It is a structural bias baked into the definition, and it points in exactly one direction. Cohort migration systematically inflates the LTH cost basis during bear phases, which mechanically compresses LTH-MVRV, which makes a sub-1 reading at the bottom of a cycle a near-mathematical certainty rather than a market signal. In the 2018 and 2022 bear markets, the process ran to completion: enough expensive coins aged into the cohort, the average cost basis climbed, price fell below it, and LTH-MVRV printed below 1. The reading was not primarily a statement about holder psychology. It was a statement about arithmetic — a cohort that had absorbed a full bear market's worth of underwater buyers.
So when the current cycle refuses to print that sub-1 reading, the interesting question is not "are holders happier?" The interesting question is "which of the two mechanisms — numerator or denominator — is responsible, and is the mechanism stable or is it an artifact?" A ratio that stays above a line because the denominator stopped inflating is telling a very different story from a ratio that stays above a line because the numerator stayed high. The Glassnode insight, as circulated, does not cleanly separate them. It reports the composite. That is the second red flag, and it is the more consequential one.
The Cohort Migration Problem: How LTH Eats Its Own Tail
Let me build on the denominator mechanism, because it deserves a full section rather than a paragraph. The dynamic rolling nature of the LTH bucket produces what is functionally a survivorship bias, and survivorship bias is one of the most reliable ways to make a metric look reassuring when the underlying reality is more ambiguous.
Consider a cohort of buyers who enter near a local top. Some panic and sell within weeks — they remain Short-Term Holders and their realized price resets to wherever they sold. Others hold. Ninety days pass. One hundred and twenty. At day 155, the survivors silently change category. They are now Long-Term Holders. Nobody voted them in. They did not pass a conviction test. They simply did not sell, and the clock ran out. Their cost basis — set near the top — is now part of the LTH aggregate.
Now scale that up across an entire bear market. Millions of coins bought at progressively higher prices across a multi-year run get held through the subsequent decline. As the decline drags on past 155 days, they all migrate. The LTH cost basis ratchets upward, not because anyone bought higher recently, but because the survivors of the previous top are now being counted. The cohort's average cost basis is a lagging indicator of the last cycle's euphoria, imported wholesale into the current cycle's accounting. This is why LTH-MVRV historically dips below 1 near bottoms: the cohort has been stuffed with expensive coins, and price has fallen below their blended cost.
This mechanism also explains a second, less discussed artifact. Because the migration is continuous, the LTH bucket is never clean. It always contains a mixture of genuinely ancient coins — the diamond hands, the lost keys, the long-dormant whales — and recently graduated coins that are only "long-term" by the technicality of having survived 155 days. The aggregate MVRV blends these two populations with no weighting disclosure. A cohort dominated by genuinely ancient, low-cost-basis coins will show a very different MVRV than a cohort that has recently absorbed a wave of expensive graduates, even if the "holder psychology" is identical. Glassnode does not, in the public material, break out this composition. That is the methodology black box, and it is not hypothetical.
I have audited systems with this exact class of hidden variable before. When I spent four months in 2017 tracing Zilliqa's Scilla contract logic and their sharding finality assumptions against their whitepaper, the failure mode was not a broken function. It was an undisclosed assumption — an edge case in transaction finality that the marketing language never acknowledged. The lesson I carried out of that exercise and into every audit since is simple and slightly unpopular: the dangerous part of a model is almost never the part it advertises. It is the part it defines silently. In Zilliqa's case it was the finality assumption. In LTH-MVRV, it is the cohort definition and the migration rule. Sharding is easy; consensus is hard — and defining a cohort is easy; disclosing what your definition does under stress is hard.
The practical consequence is that two data providers can compute "LTH-MVRV" and get materially different numbers, because the 155-day threshold is a convention, not a law, and the address-clustering method that assembles wallets into "holders" is proprietary. Different clustering means different coins in the bucket. Different thresholds mean different migration timing. Different realized-price algorithms mean different cost bases. And every one of those choices is invisible in the final chart. The single line you see on the screen is the output of at least four undisclosed methodological choices, any one of which could flip the "first since 2015" claim into a less dramatic statement. Trust no one, verify everything — including, and especially, the definitions.
The Single-Source Problem: One Feed, No Cross-Check
The report is a Glassnode insight. That is both its strength and its structural weakness, and I want to be fair about the strength before I dissect the weakness.
Glassnode sits in the top tier of on-chain data providers. Its charts are cited by institutional research desks, its metrics are embedded in trading systems, and its brand carries real weight. When Glassnode says something, the market listens, and the listening is not irrational. The firm has a track record. But track record is not the same as reproducibility, and reproducibility is the standard I hold data claims to, because I have watched enough confident analyses collapse when a second source was finally consulted.
The structural problem is that a single-source insight has a single point of failure, and the market is currently treating that single point as a consensus. The "first since 2015" framing is a strong, headline-friendly claim. It is also exactly the kind of claim that different data providers, using different methodologies, might not reproduce. CryptoQuant maintains its own cohort definitions and its own exchange-flow models; its long-term holder metrics use different threshold logic and different address heuristics. Santiment blends social and on-chain signals and would weight the cohort differently. Nansen approaches the problem from wallet labeling rather than from a rolling age threshold, which is a fundamentally different epistemic basis. None of these is inherently more correct than Glassnode. That is precisely the point. If the claim were robust, you would expect it to survive a cross-check across at least two of these methodologies. As circulated, it has not been subjected to that test.
I want to flag how the circulation itself compounds the problem. The Glassnode insight was picked up, translated, and re-compiled by media outlets — including Chinese-language crypto media — which means the version most readers encountered was already a second-hand rendering. Translation and editorial compression are lossy operations, and in a claim whose entire force depends on precise methodology, lossiness is not a minor annoyance. It is a corruption of the signal. A phrase like "first since 2015" is narratively magnetic, which is exactly why it survives compression while the caveats — the cohort definition, the single-source limitation, the numerator-denominator ambiguity — are the first things stripped out. The meme outran the methodology, and the meme is what is trading.
This is a pattern I have documented before, and it recurs because it is structural, not incidental. When I dissected the Bored Ape Yacht Club contract structure in 2021 and published an analysis of the ERC-721 limitations, the centralized metadata storage, and the gas inefficiencies baked into their specific function calls, the community reaction was polarized not because the technical findings were wrong but because the findings contradicted a narrative that had already achieved escape velocity. Once a narrative is circulating faster than the underlying data can be verified, the data stops mattering to the price and the narrative starts mattering instead. That is where we are with LTH-MVRV right now. The chart is real. The story has decoupled from it.
The Metric-Blunting Thesis: When Custody Breaks the Sample
There is a deeper structural risk that almost no one is discussing, and it is the one I would put at the top of a due-diligence memo. It concerns whether on-chain holder metrics still measure what they were designed to measure in an era of institutional custody.
The LTH-MVRV metric was built in an environment where the coins that mattered moved on-chain, sat in self-custodied wallets, and revealed their behavior through the public ledger. The entire apparatus — realized price, cohort aging, address clustering — depends on the assumption that holder behavior is legible on-chain. That assumption is weakening.
When coins move into institutional custody — ETF vehicles, custodial accounts, prime brokerage arrangements — their on-chain footprint changes. A spot Bitcoin ETF holds its coins in a custodian's wallet. Those coins may sit untouched for years, which means the age-based heuristic will classify them as Long-Term Holders. But their "behavior" is not the behavior of a long-term believer. It is the behavior of a custody arrangement. The coins do not move because the custodian does not need to move them, not because a conviction holder is refusing to sell. A growing share of the LTH cohort may soon consist of coins that are long-term only because they are custodially parked, and that systematically distorts the very signal the metric is supposed to capture.
The second-order effect is what I call metric blunting. If the marginal supply is increasingly held in custody and increasingly invisible to on-chain behavioral analysis, then the on-chain sample becomes a shrinking and increasingly unrepresentative slice of the actual market. The metric does not break loudly. It degrades quietly. It keeps printing numbers that look reasonable while measuring a smaller and smaller fraction of the coins that matter. Complexity hides risk, and the complexity here is the migration of the market's center of gravity from self-custody to institutional custody — a migration that is invisible in the metric's output but decisive for its meaning.
I saw the same class of silent degradation in the MakerDAO collateral audit I ran during DeFi Summer in 2020. The smart contracts were elegant. The migration logic was clean. But the elegance masked a structural fragility in the oracle integration — a Chainlink feed for KNC tokens that introduced a manipulation vector whose consequences would have shown up not as a code failure but as a cascade. The exploit did not fire immediately. That was the danger. A system can be simultaneously well-built and structurally fragile, and the fragility is often the part that does not show up in the happy-path output. Three major risk protocols cited that analysis, and Maker adjusted its collateral thresholds, precisely because the risk was invisible until someone went looking for it. The metric-blunting thesis is the same shape: the LTH-MVRV number looks fine, and the fine-ness may be an artifact of a sample that no longer represents the market.
The Supply-Side Read: What Is Actually Supportable
Strip away the narrative and there is a defensible core to the insight. Let me state it as carefully as I can, because I do not want to be the analyst who dismisses a real signal while hunting for a phantom.
The defensible core is this: if the long-term holder cohort is genuinely not entering loss, then the structural sell pressure that historically accompanies cohort-wide drawdowns is attenuated. The logic is straightforward. Capitulation selling is driven by holders in loss deciding to stop the pain. If the cohort is not in loss, that specific motivation is weak or absent. Supply tightens. Price finds a firmer floor. On that reading, the LTH-MVRV observation is a supply-side signal, and supply-side signals in a fixed-supply asset with a hard cap and a programmed halving schedule are not nothing.
But I want to attach three qualifiers, because the supply-side read is where the bulls and I part ways on degree rather than direction.
First, the signal is a cohort average, not a universal truth. "The LTH cohort is in profit" does not mean every long-term holder is in profit. It means the cohort's blended cost basis is below spot. Within that blend, there are certainly holders who bought near a local top and remain underwater; they are simply outvoted in the average by older, cheaper coins. A cohort average is a political statement about whose voice counts in the aggregate, and the oldest, cheapest coins always win that vote. The individual holder near the top is invisible in the headline.
Second, the supply-side logic cuts both ways, and the bearish reading is at least as coherent as the bullish one. If the long-term holder cohort is not selling into strength, that can mean conviction — or it can mean the cohort has not yet reached the price at which its members intend to sell. LTH-MVRV at the top of a cycle historically signals distribution, not accumulation: long-term holders realize profits near euphoric peaks, and their selling is one of the markers of a top. So a cohort that is comfortably in profit and not yet distributing is consistent with two very different futures. It could be a cohort that will keep holding through a continued rise. Or it could be a cohort that has simply not yet reached its target exit. The current "recovery" in LTH-MVRV, moving up from its cycle low, is the early part of a trajectory whose destination — distribution or continuation — is the actual signal. The early part tells you almost nothing about the destination.
Third, and this is the uncomfortable one: an absence of deep drawdown is not proof of a structurally healthier market. It is equally consistent with a market that has not yet completed its clearing process. The traditional bear market does something specific: it forces expensive coins into the hands of cheaper buyers, it resets cost bases, it purges leverage, and it produces the kind of washed-out cohort that can support the next accumulation phase. If this cycle has not produced that clearing — and the absence of a sub-1 LTH-MVRV reading is at least weak evidence that it has not — then the clearing may simply be deferred rather than avoided. A cohort that never got flushed is a cohort that still carries its expensive coins, and expensive coins become selling pressure the moment price approaches their cost basis. The bullish reading of "no capitulation" and the bearish reading of "no clearing" are the same observation with opposite signs, and the chart does not tell you which one is true.
The Regulatory Layer: A Commodity With a Narrative Problem
I would be remiss if I did not address the regulatory dimension, both because it bears on how this insight will be used and because the Ethereum ETF episode gave me a very specific lens on it.
On the merits, the regulatory profile here is benign to the point of being a non-issue. Bitcoin has been broadly treated as a commodity rather than a security — the CFTC has been explicit, the SEC's posture has been at minimum accommodative, and the approval of spot Bitcoin ETFs in early 2024 settled the question for practical institutional purposes. Applying a Howey analysis, Bitcoin fails the "common enterprise" prong and the "efforts of others" prong: there is no issuer, no promoter, no shared venture whose managerial efforts drive returns. The insight itself is pure data analysis. It involves no token issuance, no fundraising, no securities offering. Regulatory risk to the content is essentially zero.
But the risk migrates to the distribution layer, and that is where the Ethereum ETF critique I wrote in 2024 becomes relevant. When I analyzed the spot Ethereum ETF filings and the staking mechanisms issuers proposed, the recurring theme was that traditional finance compliance and permissionless blockchain mechanics are structurally mismatched — the framework did not adequately account for validator slashing risk, and the gap was not a detail but a category error. The same category mismatch appears here in miniature. A piece of neutral on-chain analysis becomes a compliance problem the moment it is repackaged as investment guidance, and repackaging is exactly what happens when a viral chart circulates through retail channels.
The specific exposure is jurisdictional. In markets with strict rules on investment advice and financial promotion — mainland China is the obvious example — a translated, decontextualized on-chain insight that reads as "long-term holders are profitable, therefore the asset is attractive" can brush against promotion and advisory red lines. The exposure attaches to the republisher, not to Glassnode, and not to the data. But it is real, and it is the kind of risk that a due-diligence process should flag even when the underlying content is squeaky clean. The cleanest data in the world becomes a liability when it is amplified by a channel that does not understand what it is amplifying.
Contrarian: What the Bulls Got Right
I have spent several thousand words dismantling the narrative. Let me now do the harder and more honest thing and argue the bulls' case at full strength, because a critique that cannot steelman its opposition is not a critique; it is a posture.
The bulls are right that the macro structure of this cycle is different, and they are right that the difference is not cosmetic. The 2024 approval of spot Bitcoin ETFs was a genuine regime change. For the first time, a large and growing pool of capital can access Bitcoin through regulated, institutional-grade wrappers, and that capital behaves differently from retail capital. It is slower, stickier, and far less prone to the reflexive capitulation that produced the sub-1 LTH-MVRV readings of 2018 and 2022. If the marginal buyer is an allocator rebalancing a portfolio rather than a retail trader chasing a narrative, then the depth and duration of drawdowns should compress — and a compressed drawdown is exactly what keeps the LTH cohort out of loss. On that reading, the "first since 2015" anomaly is not an artifact at all. It is the predictable output of a market whose investor base has structurally changed.
The bulls are also right that supply dynamics matter more than sentiment in a hard-capped asset. Bitcoin's issuance schedule is programmatic and its halving is not subject to negotiation. If the long-term cohort is not selling, and issuance is declining, and institutional demand is structurally new, then the supply-demand imbalance is a real and durable tailwind that does not require any particular psychological claim about holders. I have been skeptical of scalability marketing since the Zilliqa days, but I have never been skeptical of arithmetic, and the arithmetic here is not unfavorable.
And the bulls are right, finally, that the burden of proof cuts both ways. I have argued that the "no capitulation" reading is ambiguous. Fair. But ambiguity is not refutation, and the absence of a sub-1 reading is, at minimum, a data point that the traditional cycle model did not predict. When reality diverges from a model, the intellectually honest move is to consider that the model may be wrong — not merely that reality is being misread. The bulls are not required to prove the cycle has changed. I am required to consider that it might have. That is the standard I hold myself to, and I will not pretend the anomaly is meaningless just because it is over-hyped.
Where I part ways with the bulls is not on the direction of the argument. It is on the certainty. The structural-change thesis is plausible, possibly even likely. But it is being asserted on the strength of a single metric from a single source with an undisclosed methodology, circulated through lossy media channels, and consumed by an audience that is motivated to believe it. Plausible is not proven, and the gap between the two is where capital gets destroyed.
Takeaway: An Accountability Call
The LTH-MVRV insight is a real signal wrapped in an unearned conclusion. The signal is worth tracking. The conclusion — that this cycle is structurally different because long-term holders have not gone underwater — is a hypothesis, not a finding, and it is being traded as though the distinction did not exist.
Here is what a rigorous treatment would require, and what the current discourse is missing. Decompose the ratio into its numerator and denominator effects, so we can see whether the anomaly comes from price resilience or from cost-basis dynamics. Cross-check the claim against at least one independent data provider with a different methodology, so we know whether "first since 2015" survives replication. Disclose the cohort composition, so we know how much of the LTH bucket is genuinely ancient versus recently graduated. And model the metric-blunting risk, so we understand how much of the cohort is self-custodied behavior and how much is custodial parking. Until those four questions are answered, the chart is a hypothesis generator, not a signal generator.
I have been doing this long enough to know that the market rarely waits for answers. It prices the narrative, and it corrects later. That is fine — I am not here to tell anyone what to do with their capital. I am here to insist on a standard. The next time someone shows you a chart and tells you this cycle is different, ask them one question: different from what, measured how, verified against whom? If the answer is a single line from a single source with an undisclosed methodology, you have not been given a signal. You have been given a story. Verify it before you believe it — because the meme is trading, and the methodology is not.
