Adding one venue moved the measured market more than any week ever has

+28.8%

Restatement

$15.46M published, $19.92M actual

95%

Of it is Monica

$4.24M of $4.46M

+20.5%

Largest weekly move we had ever published

smaller than this restatement

$166M

Monica's book before Thursday

no public figure anywhere

On 24 August we published $15,463,527 for the week of 17–23 August and called it the largest week in our coverage. It was.

Today the same week reads $19,919,011.

Nothing about the method changed. We started publishing Monica.

The decomposition

+28.8%, and 95% of it is one venue.

The claim

The largest weekly change we have ever reported in this newsletter is +20.5% — last Monday's record. A single coverage addition just moved the same index by more than that.

Our reading: African stablecoin volume is systematically under-measured, and the size of what is missing is larger than the moves everyone spends their time explaining. Including us. We have written four issues about why a week rose 12% or fell 6%. Those analyses were correct and they were rounding error next to what we could not see.

The obvious objection

A well-informed person would say: that is not a finding, it is your coverage being incomplete. Every expansion does this. You have simply described the mechanics of adding a venue.

That objection is fair as far as it goes, and it is why this note exists rather than a press release. Two things push back on it.

First, Monica was not obscure and it was not small. $166M through 1,074,251 deposits since 2024, from a company that has publicly stated an all-time ₦500bn processing figure. It was not a startup that appeared last quarter.

Second, no one else had a number for it. Every general-purpose wallet-labelling tool returns "unknown" the moment money touches a Nigerian fintech. There was no public on-chain figure for Monica anywhere before last Thursday — not a wrong one, none. The gap was not that the market disagreed with our estimate. It was that there was no estimate.

That is the difference between an index that is imprecise and a market that is unobserved. We think this one is still closer to the second.

What follows from it

We are not going to stop reporting weekly moves — the series is the product and it only becomes useful by being unbroken. But we will be more careful about the word record, and every issue from here restates the prior week in its method statement, so a reader can always see what moved because the market moved and what moved because we could finally see it.

The honest summary of last Monday's issue is: the record was real, and it was 29% smaller than the market it was measuring.

We would rather publish that sentence than quietly reprint the number.