A rally that begins from light positioning is a different event from a rally that begins from heavy positioning, and the difference matters to anyone deciding whether to follow it. The CFTC’s most recent census makes the distinction available.
What the data shows
- Bitcoin has moved +24% since the report date, from $64,681 to $80,275. Ether has moved +31%.
- Going in, asset managers held a net +2,732 contracts in bitcoin, −86.2% from their 2024 peak.
- In ether the same category was net −3,076 contracts, close to the most negative reading in the series.
- The first release covering the move is published tomorrow. Until then, every claim about who bought is inference.
Asset managers’ net bitcoin position through the 2026 decline
Contracts, weekly. The shaded band marks the period covered by this year’s positioning series.
Source: CFTC Commitments of Traders, Traders in Financial Futures, futures only. Prices are Coinbase daily closes. Computed by Institutional Asset Journal.
Three explanations, ranked by how testable they are
The exposure moved rather than left. An allocator that held its bitcoin exposure through futures in 2024 and holds it through a spot exchange traded product in 2026 has not reduced anything. It has changed wrapper, and the futures report would show exactly the pattern above. This is the most likely partial explanation and the report cannot confirm or refute it, because the product holdings are disclosed elsewhere and on a different schedule.
The carry trade stopped paying. Institutional longs in this contract have historically been the financing leg of a cash and carry trade. When the basis narrows the trade is taken off, and both the asset manager long and the leveraged fund short shrink together. Leveraged funds are net −7,439 contracts, well below where they started the year, which is consistent with this.
Mandates tightened after a drawdown. Bitcoin closed −48.1% from its record on the report date and ether −60.3% from its own. Investment committees respond to realized drawdowns with a lag, and the lag is usually measured in quarters. This explanation is consistent with the data and is the hardest of the three to test.
The report cannot tell you why a position left. It can tell you, precisely and on the record, that it is not there.
Institutional Asset Journal
What tomorrow’s release settles
The release covering the Tuesday inside this move is published tomorrow at 15:30 Eastern. It settles one question cleanly: whether the regulated institutional category participated in the move or watched it.
It will not settle where the buying came from if the answer is that they did not. Spot exchange traded product flows, offshore perpetual futures and over the counter activity would all sit outside it. The honest position after tomorrow, in the most likely case, will be that regulated institutional futures positioning did not drive this and that the alternative explanations are not directly observable.
The series so far
This column has run since January on one dataset and one method. The findings have been consistent: institutional positioning in regulated digital asset futures fell through the first half of 2026, ether positioning turned negative and set a record, open interest fell across four of five contracts, and the number of institutions holding a reportable long remained in single digits throughout.
None of that is a forecast, and the past ten days are a useful reminder of why it should not be read as one. Positioning describes who is in the market. It does not describe who is about to arrive.
How this was measured
- Figures are taken from the CFTC’s Commitments of Traders release, Traders in Financial Futures, futures only. That report covers positions held at the close on a Tuesday and is published the following Friday at 15:30 Eastern, so every reading here is three days old on the day it appears. Net positions exclude spread holdings, which the CFTC reports separately. Notional values apply the CME contract multiplier and the Coinbase daily close on the report date. The series is rebuilt from source each week; the CFTC restates prior weeks on occasion and restatements are carried through rather than footnoted.
