Four venues now report total flow instead of the retail band
Onboard, Grey Finance, Accrue and Raenest run business and personal flow on one rail with no on-chain signal separating them. Applying a size band to a book that cannot be separated produces a number that is neither retail nor total, so we now report their total observed flow. The whole series is restated upward — the week of 24 August from $18.96M to $20.51M.
- 4 of 19
- Venues affected
- $20.51M
- Week of 24 Aug, restated
- +539% → −22.7%
- Onboard this week
- $83.89M
- August, restated
Onboard, Grey, Accrue, Raenest
published as $18.96M
the artefact the change removes
from $78.62M
Four venues in our coverage — Onboard, Grey Finance, Accrue and Raenest — run business and personal flow through the same addresses, with the same movement patterns and no separate settlement path.
We said in August that for venues like these we would not estimate a split, and would show total flow labelled as including retail. The published index was still summing the retail size band. From today it does not.
Why the band was the wrong tool
Our retail filter is a size filter. It classifies deposits by how large they are, not by who made them. On a venue where consumer and business books are genuinely separate, that works: the large tickets are the business book, and removing them leaves the consumer one.
On a venue where the two share a rail, it does not separate anything. It just removes the large half of a single, inseparable book — and what remains is neither the retail figure nor the total figure. It is an artefact of where the threshold happens to sit.
The proof is Onboard, this week
On the old basis Onboard appeared to grow 539% — $149,994 to $959,071.
Nothing of the sort happened. Onboard's total flow fell 22.7%, from $1,241,023 to $959,122. What changed was the share of that flow landing inside the band:
| Week | Total flow | Inside the band | Share |
|---|---|---|---|
| 3 Aug | $1,023,439 | $199,938 | 19.5% |
| 10 Aug | $1,798,285 | $182,968 | 10.2% |
| 17 Aug | $498,075 | $172,528 | 34.6% |
| 24 Aug | $1,241,023 | $149,994 | 12.1% |
| 31 Aug | $959,122 | $959,071 | 100.0% |
The band was producing a flat $150,000–200,000 line for a venue whose actual flow swung between $500,000 and $1.8M. The week it happened to catch everything, it printed a 539% rise. We would have published that number today.
What it does to the series
Every period moves up. The four venues contribute their full observed flow rather than a slice of it.
| Period | Published | Restated | |
|---|---|---|---|
| Week of 17 Aug | $19,919,011 | $20,476,513 | +2.8% |
| Week of 24 Aug | $18,955,618 | $20,505,846 | +8.2% |
| July | $66,703,833 | $68,689,824 | +3.0% |
| August | $78,619,725 | $83,885,218 | +6.7% |
Of the week of 24 August's $1,550,228 move, $1,505,164 is this change and $45,065 is ordinary attribution drift.
The other fifteen venues are untouched. Any figure cited from an issue published before today should be re-checked against the current series.
What this is not
It is not a claim that these four venues are more retail than we thought, and their flow should not be read as consumer volume. It is the opposite: we are saying we cannot tell, so we will publish the whole number and label it rather than publish a filtered number that implies a precision we do not have.
These venues remain excluded from every B2B total and ranking we publish, for the same reason.
Method
The retail classification engine applies a deposit-size band that removes business-sized flow. It runs on venues where consumer and business activity are separable; venues are tagged retail_only, under_review or not_separable in our coverage seed. This change stops the band running on the four not_separable venues and reports their total observed flow instead. No address attribution, no venue coverage window and no verification filter changed. Restatement figures compare the series as published on 31 August 2026 against the same series recomputed on 7 September 2026, and separate the treatment change from ordinary attribution drift.
How attribution works →Cite this
Free to cite with attribution. No embargo, no approval needed.
AfriFlux (2026). Four venues now report total flow instead of the retail band. Published 7 September 2026. https://www.afriflux.xyz/feed/not-separable-venues-report-total-flow
Reproducing a chart? Use this line: Source: AfriFlux — on-chain intelligence for Africa's stablecoin economy. afriflux.xyz