Bitcoin's largest holders sat through a 53% drawdown and came out the other side holding more than they started with. The question the tape now poses is not whether they accumulated. It is whether they accumulated too early.
The 62% share that wouldn't move
Wallets holding 100 bitcoin (BTC) or more control 61.9% of circulating supply today. They controlled 61.9% at the start of the year, and roughly 61.6% at the 2022 cycle bottom. Price has done the travelling. The metric has stayed anchored.
One convention has to be fixed before the cohort numbers mean anything. These bands measure balances sitting in wallets, denominated in native units, and coins migrate between wallets and between custodial structures constantly. A tier's net change carries redistribution and custody mechanics alongside genuine buying and selling.
Chart
(Source: CoinMetrics)
From October's $124,824 all-time high to June's $58,525 low – a 53% decline spanning two separate crashes – the aggregate whale cohort never turned into a supplier. It added a net 53,216 BTC in the first half of 2026 and 137,380 BTC since the November selloff. Supply changed hands between wallet tiers, rarely out of them. Whatever fear moved through the price, it did not move through the ownership base, and a holder base that deepens into a halving of price is not behaving like one bracing for lower.
What the giants shed, the whales absorbed
The aggregate hides a transfer of size, and reading it demands the right ruler. Wallets holding 100k BTC or more shed 104,047 BTC year-to-date, most of it in a single 99,518 BTC February drop. Read that as selling, and you most likely misread the plumbing.
These are balances in native units, and a wallet is a container, not a conviction. When one 100k+ address slips below the threshold, its entire balance reclassifies downward in a single print – which is precisely the February signature. The mega-cohort fell 99,518 in the same month as the 10k–100k band – large enough to be institutional, small enough to sit below the mega-custodian tier – rose 140,848.
The largest addresses in Bitcoin belong to exchanges and ETF custodians, and the in-kind creation and redemption now live in the US spot funds shuttle coins between custodial wallets without a sale ever crossing an exchange. The ≥100k decline is more likely to be a custody reshuffle wearing the costume of distribution. The coins did not leave. They changed rooms.
Chart
(Source: CoinMetrics)
That reshuffle inflates the band beneath it. Most of the 140,848 BTC that landed in the 10k–100k tier in February are the mega-wallet's own coins arriving one threshold down, not fresh buying. Measure supply held above 10k BTC and the distortion vanishes because a wallet slipping across the 100k line stays inside that total and leaves it untouched. The real February accumulation was 41,330 BTC. The 10k-and-up cohort added 48,604 across 2026 and 102,460 since November, roughly a third of what the band advertises but all of it the better-suited proxy for genuine accumulation.
February's bid, June's exit
The two crashes were not met the same way, and the second one flipped the roles. February was a clean smart-money bid. As spot broke to $63,495 and market value to realized value (MVRV) compressed to 1.15, whales added 20,261 BTC into the weakness while retail wallets below 100 BTC shed 7,015. The mid-tier 100–1k band then accumulated into every recovery month from February through April, buying the bounce it had bought the bottom of.
Chart
(Source: CoinMetrics)
June flipped the retail side of that trade. Wallets below 100 BTC turned aggressive buyers into the $58,525 low, absorbing 24,881 BTC, and for the first time this year the whale universe printed a net sale, down 11,672 on the month. Yet, the sale was shallow and it was local. The 10k-and-up core never joined it, adding 49,059 BTC in June, its single strongest month of accumulation all year. The leak to retail came from the mid-tier wallets thinning at the margin, not from the large accumulators changing their minds.
The 10k-and-up cohort met February with +41,330 and June with +49,059, its two heaviest bids of the year landing precisely on the two deepest breaks. It handed back a net 41,785 BTC across the months between, the selling concentrated in the March-to-April bounce, which is how roughly 90,000 BTC of crash-buying settles at 48,604 on the year. This is not a cohort holding by inertia. It is one buying weakness and selling strength, and its heaviest buying arrived at the lowest prices Bitcoin has printed this cycle.
Deeper low that drew less
The historical comparison inverts the intuition entirely. The 2022 bottom was the real thing on every valuation axis; MVRV at 0.75, spot at $15,758 trading 25% beneath the aggregate cost basis, the market submerged below realized price for 179 days. Into that generational low, whales sold. The ≥100 cohort shed 251,493 BTC in the six months after the November 2022 print, post-FTX deleveraging forcing size out of the biggest wallets regardless of how cheap the coin had become.
Chart
(Source: CoinMetrics)
Of that, 2026 is the photographic negative. MVRV bottomed at 1.10 in June, spot never once closed below realized price, and the market spent zero days under an MVRV of 1.10 across the entire year. By the single metric that defined the last bottom, this one hardly registers as one. And yet the whale universe bought it, absorbing 137,380 BTC in the decline since November against the 251,493 it distributed after 2022's capitulation. A deeper low drew persistent selling. A shallower one drew a persistent bid.
The paradox is the thesis. Whales are buying a bottom markedly less cheap than the one they spent months selling. Either their cost-basis discipline has collapsed, or the thing that made 2022 cheap, a forced-seller overhang with no structural bid beneath it, has been engineered out of the market.
Line at realized price
Two readings survive contact with the data, and they diverge on one number. The constructive case holds that the ETF era has drained the forced-seller reservoir that manufactured 2022's capitulation, and that with realized price anchored near $53,000 and a standing bid from fund flows and the 10k–100k cohort, Bitcoin no longer has to trade on a cost basis to clear. On that logic, the $58,525 June low is the floor, printed a deliberate 10% above the aggregate entry, and the accumulation is neither early nor brave, merely correct.
The cautious case notes that MVRV at 1.20 is a mid-range, not a low, and that every prior cycle demanded compression towards or beneath 1.0 before the durable bottom formed. If that history still binds, the whale bid here is front-running a low that has not yet arrived, and the $53,000 realized-price line, the level where MVRV crosses 1.0, is the test the market has managed to avoid all year.
The whales have already voted, and they have voted in size. Whether that vote proves early or right is decided at a realized price, and Bitcoin has not yet been forced to visit it.
