When the algo breaks, the axiom remains. And this week the algo broke on a number so clean it should make every serious analyst nervous: 11,000 BTC. That is the weekly sell volume attributed to short-term holders, down roughly 66% from a peak of 32,000 BTC. The wire copy framed it as selling pressure collapsing. Headlines called it supply exhaustion. Nobody asked the only question that matters โ down from what, measured how, against which base?
I have spent the better part of a decade watching clean numbers get dressed up as conclusions. So let me be direct about what this data point actually is, and what it is not. It is a single flow metric from an unnamed on-chain analytics provider, published without cross-validation, without a disclosed time frame, and without a definition of the very cohort it claims to measure. That is not a bearish take. It is an accounting take. A number you cannot source is a number you cannot trade.
Context: What the STH Label Actually Encodes
To understand why this metric is so seductive โ and so fragile โ you have to understand what 'short-term holder' means in practice. In on-chain analytics, the industry convention is that an STH is an address whose coins have moved within the last 155 days. That threshold is not physics. It is a heuristic that Glassnode, CryptoQuant, and their imitators have broadly converged on because 155 days, roughly five months, maps reasonably well onto the behavioral boundary between traders and holders.
The classification is not observed. It is inferred. And the inference rests on two technical assumptions that almost nobody publishes alongside the number: first, that exchange address clustering is accurate enough to attribute flows to the right entity; second, that the time-bucket algorithm assigns coin age consistently. When those two assumptions hold, the STH label is a useful lens. When they do not, you are reading an artifact of someone's methodology and calling it a market signal.
Here is the part the wire copy buried. The 'sell volume' this metric describes is not realized selling. It is an exchange inflow subset โ coins moving toward venues, not coins being sold. Coins arriving at an exchange are potential sell pressure, not confirmed sell pressure. The gap between those two states is where most retail narratives are born and where most retail capital dies.
Now layer the supply base underneath it. Since the April 2024 halving, the network issues 3.125 BTC per block, roughly 144 blocks a day. That is about 450 BTC per day, or roughly 3,150 BTC per week of fresh miner issuance. Hold that number, because it is the yardstick every 'supply shock' claim has to clear.
Core: The Arithmetic Behind the 66% Headline
Start with the ratio that matters. 11,000 BTC per week of STH selling against 3,150 BTC per week of new issuance means short-term holders are still moving roughly 3.5 times the entire miner supply to market every single week. That is the 'exhausted' reading. What looks like a drought is still a flood, just a smaller one.
I ran a similar stress test in 2020, when I was a junior analyst tracking Uniswap and Curve flows. My peers were transfixed by APYs; I was transfixed by the correlation between stablecoin de-pegging risk and gas spikes. I published a thread arguing that DeFi yield was largely illusory, funded by retail liquidity rather than organic revenue, and I calculated that if Bitcoin dominance fell below 30%, the whole structure would face a liquidity crunch. Two months later the market corrected and the thesis held. The lesson I carried forward was not that I was right. It was that protocol-level health is downstream of macro liquidity, and single-metric narratives are almost always downstream of someone's incentives.
The second piece of arithmetic is uglier, and it is the thing I most want you to take from this article. The STH cohort is defined by a rolling time window. Every day, addresses whose coins cross the 155-day line mechanically 'graduate' out of the short-term bucket and into the long-term one. That means the STH pool shrinks on its own, without a single behavioral change from any holder. A falling STH sell volume can be the arithmetic residue of cohort migration rather than evidence that anyone stopped selling. If you do not control for STH total supply โ and this report does not โ you cannot distinguish 'holders locking up conviction' from 'the denominator quietly walking out the door.'
This is the structural defect that makes the metric dangerous in isolation. Suppose STH address count and aggregate holdings are both falling. Then coins are graduating into long-term hands, supply is genuinely locking, and the bullish reading has legs. Suppose STH holdings are flat while sell volume drops. Then you are watching people sit on their hands, which is neutral at best. The same headline number supports both readings. The report gives us neither.
What the data does tell us, if we read it honestly, is that the selling we are seeing is happening in profit. STH coins moving at a gain implies the cohort's average cost basis sits below spot โ a market where profit-taking dominates, not capitulation. That is a materially different psychological state from a panic flush. Profit-taking is orderly. It is also, historically, what happens in the middle of a trend, not the end of one. Capitulation is loud and brief. Distribution is quiet and patient. This number looks like the second.
The Blind Spots Nobody Priced
Let me be precise about the causal trap here, because it is where most people will lose money. A decline in STH selling has at least four explanations, and the wire copy picked exactly one:
- Sell pressure exhausted โ bullish.
- Distribution already complete โ the sellers finished, the bag is in stronger hands, and the move is done. Neutral to bearish.
- Coins moved to cold storage and went idle โ neutral.
- Cohort migration โ the STH pool shrank mechanically. Neutral, and possibly meaningless.
Only one of those four is bullish, and the source picked it without showing its work. That is narrative selection, not analysis. When I audited token models in 2017 โ after a poorly audited privacy coin I held rug-pulled within days โ I learned that the failure was never in the code alone. It was in the story people told themselves about the code. The 11,000 BTC figure is that same story in a new costume.
The deeper blind spot is what a single flow metric cannot see. A defensible on-chain read requires STH-SOPR, which measures the profit ratio on coins actually spent; STH-MVRV, which tells you whether the cohort is in aggregate profit; exchange netflow, to distinguish inflow from outflow; and long-term holder behavior, to see whether conviction is genuinely rising. This report gives us one of those four. One data point is not a thesis. It is a hypothesis wearing a thesis's clothes.
And there is a wrinkle that the ETF era introduced, one I flagged when I analyzed custodial risk in multi-sig wallets after the January 2024 spot approvals. Some of what registers as 'STH selling' may now originate from ETF creation and redemption mechanics rather than from independent on-chain addresses. The instruments have changed. The labels have not caught up. On-chain metrics can undercount or misattribute real sell pressure in a market where a meaningful share of flow runs through custodians that never touch a labeled address.
Contrarian: The Weakness Is the Point
Here is where I part ways with both the bulls and the bears. The bulls will cite this number as dry powder for a squeeze. The bears will dismiss it as noise. Both miss that the most important thing about this data point is not its direction โ it is its low information density.
Think about the base. Bitcoin has no team allocation, no unlock cliff, no vesting schedule, no foundation wallet that can dump on you. It is, structurally, the cleanest supply profile in the asset class โ roughly 19.8 million of 21 million coins already mined, with no insider overhang and no governance attack surface. That cleanliness is exactly why a metric like this gets weaponized. When an asset has no fundamental drama to trade on, the market manufactures drama out of flow data. Skepticism is the highest form of due diligence, and it is never more necessary than when the underlying story is this thin.
Here is the counter-intuitive read. If this were a genuine supply shock, we would see it in the cross-confirmations: sustained exchange netflow outflows, STH aggregate supply falling, funding rates flipping positive while price holds, and ETF creation running net positive. We do not have any of those in this report. What we have is a single number, framed against a peak, using the strongest available comparative rhetoric. 'Down 66% from the peak' is not analysis. It is a framing device engineered to feel like a conclusion.
The macro layer makes this sharper, not softer. In a bull market, flow metrics get read as confirmation of a trend that liquidity is already driving. That is backwards. Liquidity sets the tide; on-chain flow describes the foam. If global M2 is expanding and rates are easing, this number is a curiosity. If liquidity is tightening underneath a euphoric tape, the same number becomes a warning that the sellers who left were simply the first to read the water level. The data is synchronous at best. It confirms. It does not lead.
Takeaway: What to Watch Instead
Do not trade the headline. Trade the convergence. The only version of this signal worth acting on is the one where four things align: STH aggregate supply falling, STH-MVRV above one with SOPR cooling, exchange netflow persistently negative, and spot ETF creation net positive over the same window. That is supply tightening against demand expanding โ the real squeeze. Anything less is a single number wearing a costume.
So watch the denominator, not the numerator. If STH holdings shrink while sell volume falls, conviction is real. If they hold flat, you are watching indecision dressed as strength. And if price refuses to rally on a 'bullish' flow print, you have your answer: the sellers did not leave because they believed. They left because they were finished.
When the algo breaks, the axiom remains. The axiom here is simple โ you cannot measure pressure you cannot source, and you cannot trade a story you cannot cross-check. The next two weeks will tell you which of the four explanations was true. Watch the water, not the foam.