Bitcoin Supply Shock Eases to 2.6x, Last Week Just 0.6x
How much of the new bitcoin supply did ETFs absorb this week?
Over the 20 trading sessions from 8 September to 5 October, US spot Bitcoin ETFs took in a net 23,231 BTC while miners created 9,000 BTC. That is a bitcoin supply shock reading of 2.58x: the funds absorbed roughly two and a half coins for every one that entered existence. The reading has fallen for a second straight week, from 3.78x a week ago, and the most recent five sessions managed only 0.63x — ETF demand slipped below bitcoin daily issuance for the first time since the start of September.
Both halves of that matter. A 20-session window still comfortably above 1.0 means the structural squeeze is intact. A trailing week below 1.0 means it is not being renewed right now.
Why the bitcoin supply shock reading fell 1.2 turns in a week
The denominator never moves. Bitcoin's issuance is a fixed step function — 450 BTC per day since the April 2024 halving — so every change in this ratio comes from the numerator. Last Tuesday's window held 34,005 BTC of absorbed supply. This week's holds 23,231 BTC. The entire 1.20-turn decline is 10,774 BTC of absorption that left the window.
Very little of that is new selling. It is mostly the arithmetic of what the window covers. The exceptional run of late September — 11,528 BTC on 21 September alone, 25.6x that day's issuance, the largest single session of 2026 — is still inside the window, but the sessions that have joined it since are small. The 20-session net dollar figure behind this is $2,080.5M, converted to BTC at each day's own close rather than at a single average price, which is why the coin count and the dollar count can drift apart when price moves.
The window is also less one-sided than the headline suggests. Of its 20 sessions, 12 were net positive and 8 were net negative. This is not twenty days of steady buying; it is a handful of very large buying days carrying a lot of flat and negative ones.
The last five sessions bought less bitcoin than miners produced
Narrow the window and the picture changes sharply:
- Last 20 sessions: 23,231 BTC absorbed against 9,000 mined — 2.58x
- Last 10 sessions: 18,055 BTC against 4,500 mined — 4.01x
- Last 5 sessions: 1,424 BTC against 2,250 mined — 0.63x
The 10-session figure is higher than the 20-session figure because the late-September burst sits inside it. The 5-session figure is the one to watch: from 29 September to 5 October the funds took in less than two thirds of what miners produced over the same stretch. Two of those five sessions were net redemptions, including 5 October at -1,047 BTC.
That is a real deceleration, and it is consistent with what Monday's flow article found in dollar terms — inflows increasingly dependent on a narrow set of issuers. Converted to coins, a narrow bid is a smaller bid.
Where 2.58x sits against 2026's own absorption range
Context is where this metric earns its keep, because the absolute number means little without the distribution behind it. Across the 682 rolling 20-session readings in the series, 2.58x sits at the 54th percentile — almost exactly mid-range. Measured only against post-halving sessions, it is the 56th percentile, against a post-halving median of 1.73x.
Against 2026 specifically it looks much stronger. This year's 190 rolling readings have a median of just 0.537x, and only 44% of them cleared 1.0 at all. The year's range runs from -8.62x in June to 5.25x in early May. On that scale, 2.58x is a good reading in a mediocre year — 55 of this year's 190 readings were higher, but the typical 2026 session has seen ETFs absorb barely half of new supply.
So the honest framing is neither "supply shock over" nor "supply shock accelerating". It is a reading that has come off a strong September but remains well above where most of 2026 has traded.
Next week's reading should rise even if nobody buys
Here is the mechanical quirk worth flagging before it gets misread. The five oldest sessions in the current window — 8 to 14 September — are a net -3,949 BTC. Those sessions roll off over the coming week.
If the next five sessions were completely flat, with zero net flow, the 20-session absorption ratio would still climb from 2.58x to roughly 3.02x, purely because a block of redemptions dropped out of the comparison. If the headline ratio rises next Tuesday, that will be a base effect first and a demand signal second, and anyone quoting the number without checking which sessions left the window will draw the wrong conclusion from it. This is the main reason the site's chart on the institutional signal dashboard plots the whole rolling series rather than a single current figure.
What bitcoin daily issuance of 450 BTC does to the arithmetic
The halving is doing more work in this ratio than demand is, and it is easy to miss. Take this week's 23,231 BTC of absorbed supply and measure it against the pre-halving issuance rate of 900 BTC per day. The same demand, against the old supply schedule, would be 18,000 BTC of issuance and a ratio of just 1.29x — half of what it reads today.
Nothing about investor behaviour would have changed in that comparison. Only the denominator. The series itself straddles the step: 72 of its sessions were mined at 900 BTC per day and 629 at 450, with the first 450-per-day session on 22 April 2024. Every post-halving reading is structurally twice as high as the identical dollar demand would have produced before.
This is the core of the bitcoin scarcity argument, and also its most over-claimed part. The halving mechanically doubles this ratio for unchanged demand. It does not create demand. A reading of 2.58x today and 2.58x in early 2024 would describe two quite different worlds of buying pressure.
ETFs now hold 686,247 BTC, or 1,525 days of issuance
Cumulatively, since the funds began trading in January 2024, US spot Bitcoin ETFs hold a net 686,247 BTC. Two ways to size that:
- At the current 450 BTC per day, that stock equals 1,525 days of issuance — more than four years of future mining output, already absorbed.
- Across the series' own span, miners produced 347,850 BTC. The funds hold 1.97x everything mined since the first ETF session.
Those are stock figures, not flow figures, and they do not go away when a week disappoints. A ratio below 1.0 for five sessions removes almost nothing from a position of this size. That is the asymmetry people reach for when they talk about an ETF-driven bitcoin supply shock: the flow is volatile, the stock is sticky.
Strategy added 334 BTC, and the week still came up short
The ETF channel is not the only corporate bid competing with miners for coins. Strategy's 8-K filed 5 October reports 334 BTC acquired for $28.7M at an average of $85,928, taking total holdings to 848,000 BTC. Across the same 20 sessions as the absorption window, Strategy's filings show 2,949 BTC acquired and no disposals — about 12.7% on top of the ETF figure, lifting combined absorption to 26,180 BTC and the ratio to 2.91x.
It does not rescue the trailing week. Adding Strategy's 334 BTC to the funds' 1,424 BTC gives 1,758 BTC against 2,250 BTC mined over those five sessions — 0.78x, still short of issuance. Thursday's treasury coverage takes the financing side of that purchase apart properly; here it is only a second supply sink, and this week a small one.
What this bitcoin scarcity measure cannot see
Every number above is ETF flow plus, where stated, Strategy's filings. The limits are worth stating plainly, because this is a demand proxy and not a supply ledger:
- It counts one channel. OTC desks, exchange accumulation, sovereign buyers, private funds and retail self-custody are all invisible to it. A reading of 0.63x does not mean the market absorbed only 63% of new supply; it means the ETFs did.
- A negative reading is redemptions, not miner oversupply. When the ratio went to -8.62x in June, that was investors handing coins back to the funds. Miners cannot produce more than 450 BTC per day no matter what the ratio does.
- Miners are not the only sellers. Issuance is the cleanest supply benchmark because it is fixed and knowable, but long-dormant coins also come to market, and this ratio does not see them.
- Flow is converted at each day's close. A large flow on a day of sharp price movement converts to a coin count sensitive to that close.
- It is not a price forecast. Absorption above 1.0 has coincided with falling prices and below 1.0 with rising ones. Bitcoin closed at $85,738 on 5 October, above the window's $81,457 average close, in a week when ETF demand ran below issuance. Treat the ratio as a measure of how much of the new supply one large buyer group is taking, nothing more.
Within the site's institutional composite, supply absorption still scores 81.1 out of 100 and contributes the largest positive share of the overall reading — a reminder that on a 20-session view this remains the strongest part of the demand picture, even as the last week of it softened. The full series and the week's dollar detail are on the ETF flow dashboard, and last Tuesday's supply absorption read covers the 3.78x peak this week stepped down from.