$90K Is the Nov 2025-Jan 2026 Distribution Shelf, Not a Ceiling: BTC Spent 74 Days at $85-100K and Last Closed Above $90K on 19 Jan 2026; CryptoQuant's 365-DMA 'Bull Confirmed' Signal Is Half a Falling Average ($91.5K in June to $80.4K Now) and Only 2 of 3 Prior Reclaims Since 2015 Held
What the parent card says, checked. The Cointelegraph piece (23 Sept 2026, William Suberg) relays CryptoQuant's weekly report: price reclaimed the 365-day moving average at $80,500; the 1-to-3-month 'trader realized price' is $64,300; its upper band (+40%) is $90,300 and coincides with an $88K-$90K on-chain supply cluster, so $90K is a 'natural pause point within an uptrend, not a reversal'. Every one of those numbers reproduces from public data. Coin Metrics community daily closes (fetched 23 Sept) give a 365-day simple average of $80,435 on 22 Sept, and the mean close of 24 June-23 Aug (the 1-3 month window) is $64,320. The MVRV claim also checks out: the 2026 low was $58,525 on 30 June with MVRV 1.10, the highest MVRV cycle-low in Coin Metrics' series apart from the 2021 mid-cycle dip (1.54) and 2024 (1.71); 2018, 2019, 2020, 2022 and 2023 all printed below 1.0 (0.69, 0.76, 0.88, 0.75, 0.84). The parent is accurate. What it does not tell you is where $90K comes from, how soft the 'confirmation' signal is, and what the base rate looks like.
Where the $88-91K wall comes from. Coin Metrics daily closes show BTC spent 74 of the 92 days from 1 Nov 2025 to 31 Jan 2026 between $85K and $100K: 19 days at $88-90K, 20 days at $90-92K, 12 at $92-95K, 6 at $95-100K, all in that window and nowhere else in the past year. That is the shelf where the market distributed on the way down from the $124,824 6 Oct 2025 peak (the drawdown to the 30 June low was 53.1% from the ATH and 39.7% from the $97,044 14 Jan 2026 high). The last daily close above $90K was $92,526 on 19 Jan 2026, eight months ago. So the 'supply cluster' is not an abstraction: it is the cost basis of everyone who bought in Q4 2025 and is now, for the first time since January, within 5% of breakeven. CryptoQuant's own framing has been shifting up with price: on 12 Sept (The Block, Julio Moreno) the same report set the 365-DMA at $81,700, the trader-realized upper band at $88,700, a $77,100-$80,200 zone where 539,000 BTC had been sold in 30 days, and a $62-65K zone where long-term holders accumulated 476,000 BTC; eleven days later the band is $90,300 and FXStreet's 23 Sept write-up quotes the cluster as $88-91K with the $76-81K zone 'cleared'. The bands are trailing indicators of a rising 1-3 month cost basis, which is why they move.
The 365-DMA 'confirmation' is half a falling average. This is the point the headline misses. Price crossed above the 365-DMA on 18 Sept ($80,944 vs $80,875), but the average itself has been falling about $2,800 a month: $91,526 on 24 June, $87,398 on 24 July, $83,216 on 23 Aug, $80,435 on 22 Sept. Of the gap that closed since June, price rose $27.7K (+47.3% from the low) and the average fell $11.1K. The closes rolling off next are the $109K-$124K prints of Sept-Dec 2025 (the next 30 days of roll-off average $114,331; the next 90 average $102,226), so even at a flat $85,450 spot the 365-DMA falls to about $78,060 in 30 days and $76,300 in 90. Being above it in Q4 2026 requires only that BTC not fall 10%; it is a low bar, not a verdict.
Base rate for the reclaim. Since 2015 there have been three prior cases of BTC closing back above its 365-DMA after at least 60 days below it (Coin Metrics daily): 3 May 2019 ($5,658, after 276 days below; +54.5% in 30 days, +83.8% in 90, no close below entry within 90 days), 27 March 2022 ($46,777, after 84 days; -18.6% in 30 days, -54.1% in 90, the reclaim failed), and 14 March 2023 ($24,769, after 343 days; +22.6% in 30, +4.6% in 90, worst close -1.5%). Two of three held. The failure had MVRV at 1.91; the successes were at 1.30 and 1.25; today's reclaim came at 1.52 (18 Sept) and sits at 1.61 (22 Sept), between the two regimes. That is the honest read: the signal has a 2-of-3 record and current valuation is closer to the 2022 failure than the 2019 and 2023 successes.
The demand side is real but concentrated. US spot ETFs took $999M on 21 Sept (The Block data, 22 Sept: IBIT $381.4M, ARKB $289.1M, FBTC $238.8M, Grayscale Mini $61.7M, BITB $21.6M) and Cointelegraph reports $1.7B over the two days to 22 Sept, the largest one-day print since Oct 2025. Yesterday's card on this feed showed the basis trade explains at most about $3.4B of the streak, so most of it is directional. Ki Young Ju's 22 Sept X post (crypto.news, 23 Sept) argues the same institutional ownership caps the upside: he expects 3-5x for this cycle, not 10x, notes realized cap grew $697B for a 689% gain last cycle versus $2.7B for 55,000% in the early years, and puts the capital needed for another parabola above $1T.
What to watch. A daily close above $92,526 (the 19 Jan high-water mark) would be the first since January and would put the entire Q4 2025 shelf in profit; a rejection at $88-91K with ETF flows still positive is the base case CryptoQuant describes and is consistent with the $85-90K dealer gamma wall into the 25 Sept $16B Deribit expiry covered yesterday. The 365-DMA will not be the tell either way, because it falls to $78K by late October on its own.
Sources (6)
AI Research
Key Takeaway
CryptoQuant's numbers reproduce from Coin Metrics data, but $88-91K is the cost basis of the 74 days BTC spent at $85-100K in Nov 2025-Jan 2026 (last close above $90K: 19 Jan), the 365-DMA 'confirmation' is mostly the average falling from $91.5K to $80.4K since June, and prior reclaims held only 2 of 3 times.
What the parent card says, checked. The Cointelegraph piece (23 Sept 2026, William Suberg) relays CryptoQuant's weekly report: price reclaimed the 365-day moving average at $80,500; the 1-to-3-month 'trader realized price' is $64,300; its upper band (+40%) is $90,300 and coincides with an $88K-$90K on-chain supply cluster, so $90K is a 'natural pause point within an uptrend, not a reversal'. Every one of those numbers reproduces from public data. Coin Metrics community daily closes (fetched 23 Sept) give a 365-day simple average of $80,435 on 22 Sept, and the mean close of 24 June-23 Aug (the 1-3 month window) is $64,320. The MVRV claim also checks out: the 2026 low was $58,525 on 30 June with MVRV 1.10, the highest MVRV cycle-low in Coin Metrics' series apart from the 2021 mid-cycle dip (1.54) and 2024 (1.71); 2018, 2019, 2020, 2022 and 2023 all printed below 1.0 (0.69, 0.76, 0.88, 0.75, 0.84). The parent is accurate. What it does not tell you is where $90K comes from, how soft the 'confirmation' signal is, and what the base rate looks like.
Where the $88-91K wall comes from. Coin Metrics daily closes show BTC spent 74 of the 92 days from 1 Nov 2025 to 31 Jan 2026 between $85K and $100K: 19 days at $88-90K, 20 days at $90-92K, 12 at $92-95K, 6 at $95-100K, all in that window and nowhere else in the past year. That is the shelf where the market distributed on the way down from the $124,824 6 Oct 2025 peak (the drawdown to the 30 June low was 53.1% from the ATH and 39.7% from the $97,044 14 Jan 2026 high). The last daily close above $90K was $92,526 on 19 Jan 2026, eight months ago. So the 'supply cluster' is not an abstraction: it is the cost basis of everyone who bought in Q4 2025 and is now, for the first time since January, within 5% of breakeven. CryptoQuant's own framing has been shifting up with price: on 12 Sept (The Block, Julio Moreno) the same report set the 365-DMA at $81,700, the trader-realized upper band at $88,700, a $77,100-$80,200 zone where 539,000 BTC had been sold in 30 days, and a $62-65K zone where long-term holders accumulated 476,000 BTC; eleven days later the band is $90,300 and FXStreet's 23 Sept write-up quotes the cluster as $88-91K with the $76-81K zone 'cleared'. The bands are trailing indicators of a rising 1-3 month cost basis, which is why they move.
The 365-DMA 'confirmation' is half a falling average. This is the point the headline misses. Price crossed above the 365-DMA on 18 Sept ($80,944 vs $80,875), but the average itself has been falling about $2,800 a month: $91,526 on 24 June, $87,398 on 24 July, $83,216 on 23 Aug, $80,435 on 22 Sept. Of the gap that closed since June, price rose $27.7K (+47.3% from the low) and the average fell $11.1K. The closes rolling off next are the $109K-$124K prints of Sept-Dec 2025 (the next 30 days of roll-off average $114,331; the next 90 average $102,226), so even at a flat $85,450 spot the 365-DMA falls to about $78,060 in 30 days and $76,300 in 90. Being above it in Q4 2026 requires only that BTC not fall 10%; it is a low bar, not a verdict.
Base rate for the reclaim. Since 2015 there have been three prior cases of BTC closing back above its 365-DMA after at least 60 days below it (Coin Metrics daily): 3 May 2019 ($5,658, after 276 days below; +54.5% in 30 days, +83.8% in 90, no close below entry within 90 days), 27 March 2022 ($46,777, after 84 days; -18.6% in 30 days, -54.1% in 90, the reclaim failed), and 14 March 2023 ($24,769, after 343 days; +22.6% in 30, +4.6% in 90, worst close -1.5%). Two of three held. The failure had MVRV at 1.91; the successes were at 1.30 and 1.25; today's reclaim came at 1.52 (18 Sept) and sits at 1.61 (22 Sept), between the two regimes. That is the honest read: the signal has a 2-of-3 record and current valuation is closer to the 2022 failure than the 2019 and 2023 successes.
The demand side is real but concentrated. US spot ETFs took $999M on 21 Sept (The Block data, 22 Sept: IBIT $381.4M, ARKB $289.1M, FBTC $238.8M, Grayscale Mini $61.7M, BITB $21.6M) and Cointelegraph reports $1.7B over the two days to 22 Sept, the largest one-day print since Oct 2025. Yesterday's card on this feed showed the basis trade explains at most about $3.4B of the streak, so most of it is directional. Ki Young Ju's 22 Sept X post (crypto.news, 23 Sept) argues the same institutional ownership caps the upside: he expects 3-5x for this cycle, not 10x, notes realized cap grew $697B for a 689% gain last cycle versus $2.7B for 55,000% in the early years, and puts the capital needed for another parabola above $1T.
What to watch. A daily close above $92,526 (the 19 Jan high-water mark) would be the first since January and would put the entire Q4 2025 shelf in profit; a rejection at $88-91K with ETF flows still positive is the base case CryptoQuant describes and is consistent with the $85-90K dealer gamma wall into the 25 Sept $16B Deribit expiry covered yesterday. The 365-DMA will not be the tell either way, because it falls to $78K by late October on its own.