Anatomy of a Drawdown: What Resolution Reveals
Old hands sold the late-June low on both chains, at a loss and in profit alike. What absorbed the selling was not what the dashboards showed. Ethereum’s bid was the staking lock. Bitcoin’s whale signal, put under the same lens, refused to produce a sign at all. This is the drawdown, measured on the resolved ledger.
01The story that does not survive
Read naively, the drawdown told a story of divergent conviction. Ethereum’s large holders appeared to buy the fall: a headline +6.2M ETH flowed into wallets above 10k ETH between December 2025 and June 2026. Bitcoin’s large holders appeared to sell it. Two chains, two opposite bets, one tidy narrative.
Resolved, the divergence dissolves on both chains, for two different reasons. Ethereum’s buyer was a protocol. Bitcoin’s seller was, in large part, plumbing, and what remains once the plumbing is removed does not add up to a direction. What stands afterward is quieter and better supported: long-term holders capitulated on both chains at once, and the only price-insensitive absorber in the episode was a mechanism, not conviction.
02Ethereum: the buyer was the staking queue
Quarantine protocol contracts, on the principle that a contract is never a beneficial holder, and the cohort above 10k ETH splits cleanly.
| DEC 2025 → JUN 2026 | HELD ETH | CHANGE |
|---|---|---|
| Genuine large holders | 50.548M → 45.767M | −4.78M |
| Staking sink (ETH2 Beacon) | 84.425M → 95.609M | +11.18M |
| Naive “all large” — the +6.2M read | 137.17M → 143.40M | +6.23M |
The published +6.2M of whale accumulation decomposes into genuine holders at −4.78M and the staking sink at +11.18M, with exchange entities a rounding error at −0.18M. Genuine large holders were net sellers into the fall. The marginal buyer was the deposit queue: structural, price-insensitive, and unable to sell what it takes in.
The reattribution rests almost entirely on one unarguable classification. The ETH2 Beacon Deposit Contract alone is 90.9% of all sink mass, so the sign does not hinge on any judgment call at the margins. This is a derived result rather than an estimate: protocol sinks are quarantined at birth, and the regrouping conserves mass exactly.
03Bitcoin: the direction does not resolve
Bitcoin offers no equally clean cut, and the honest finding is that no direction can be established for its large holders in this window. The same June flow reads three ways depending on construction. Raw per-UTXO nets to −13,273 BTC, a change-inflated floor. Raw entity-clustered stock reads −406,290 BTC, diffuse consolidation churn, and notably not concentrated at exchanges, which added coins over the month. A custodial-quarantined economic estimate lands positive but cannot be reproduced from persisted entity labels. Across constructions the answer spans roughly 420,000 BTC.
04What survives: convergent capitulation
None of the above touches the flow-side capitulation measures. They are computed on classified movement, independent of the whale-side correction, and they stand intact.
The through-line is convergent long-term-holder capitulation. Losers exited, and some winners exited with them. The divergent-conviction story was measurement noise riding on top of it. The drama was in the artifacts, not the holders.
05Where the market sits
By exact cost-basis interpolation, about 58% of ETH supply sat underwater at a spot near $1,900. The regime read under our v1.3 model is ETH −0.779, fear-accumulation, as of 2026-06-30, and BTC −0.009, neutral-expansion, as of 2026-07-08, Bitcoin having eased out of fear as its price recovered. Those scores are that model’s read on those dates. A later revision supersedes them; the live product carries the current read.
On the 58%: the exact figure interpolates the cost-basis band that straddles spot. A coarser convention counts the entire straddling band as underwater and yields roughly 78%. The interpolated 58% is the honest number.
06Addendum: the 2021 vintage, resolved
The cohort that bought the 2021 top still nominally holds 11.5M ETH, down from 67.0M at the peak. Resolve the remainder and the “survivors holding firm” story thins further: 6.65M, about 58% of it, sits locked in the Beacon staking contract, 0.28M sits at exchanges and other entities, and the genuine free-float survivors hold 4.57M. The actionable base is a fraction of the raw number.
07What this is, and what it is not
We are not calling a bottom. If the market revisits the lows, the ledger will show who blinks, on the resolved base this time. What is claimable, with stamps: old-coin capitulation above the 90th percentile of its history coincided with what has so far been the low of this drawdown; the whale-side comfort statistic did not survive resolution; and every figure above is a dated, reconciled fact of the ledger, or is labeled exploratory where it is not.