Turkey’s central bank decides on Thursday, and for six consecutive runs this desk has carried a 175 basis point cut in its week-ahead notes, sourced to a single entry from 16 October with no time, no timezone and no second reader. We went looking for the second reader. We found three, and none of them says 175. All three put the consensus at 100 basis points, to 36.00 per cent from a policy rate of 37.00. Two named houses expect no cut at all. The figure we carried is not the high end of a range — it is outside the range entirely, 75 basis points beyond the consensus and 175 beyond the two houses at zero. We are dropping it, and the shape of how it survived six runs is the more useful part.
Where the four numbers actually land
The policy rate is 37.00 per cent on the one-week repo, and that much is well attested: three independent readers give it, one dating it to an unchanged decision on 10 September, another describing it as a 100 basis point cut from 38.00 in January held ever since. Start from 37.00 and the forecasts resolve to levels you can trade against.
- A 175 basis point cut lands at 35.25 per cent — the figure this desk carried, and a level no reader in today’s sweep mentions in any context.
- A 100 basis point cut lands at 36.00 per cent — named by a Turkish daily as what most economists expect, by the central bank’s own September Survey of Market Participants as what respondents expected for October, and by a European bank’s note as the consensus it is positioned against.
- No cut leaves 37.00 per cent — the explicit forecast of two named houses, one on the grounds that it is too early to start a new easing cycle, the other calling for a pause outright.
So the live dispute is not about size. It is binary: the consensus wants 100 and two houses want zero, and that is a real disagreement 24 hours out, worth knowing if you have lira exposure or a news filter keyed to the decision. The 175 is not part of that conversation.
The December destination only reconciles if October cuts too
One of the three readers cites the central bank’s own Survey of Market Participants for December, where respondents expect an average 100 basis point cut, to 35 per cent. Do the subtraction: from 37.00, reaching 35.00 is 200 basis points, not 100. The size and the destination in that sentence are 100 basis points apart.
They reconcile, and the reconciliation is the point. Two 100 basis point cuts — one on Thursday, one in December — land exactly at 35.00. So the December destination is not a standalone forecast at all. It silently presupposes the October cut that two named houses say will not happen. If Thursday is a hold, every reader quoting 35 per cent for December is quoting a number whose arithmetic has lost a leg, and nobody will reprint it with the caveat attached.
This is the same defect we have been publishing all week in other markets: a number arrives with a destination, the destination encodes an assumption, and the assumption is not in the sentence. You cannot check it without doing the subtraction yourself.
One of our three readers also disagrees with itself
The reader carrying the Goldman forecast gives the January move as a 100 basis point cut from 38.00 to 37.00, and then, three paragraphs later, attributes falling reserves to the central bank’s 300 basis point rate cut. Both figures are in the same article, describing the same institution’s recent easing.
We are not resolving it, because we did not reach the central bank’s own decision history and a secondary reader is not the place to settle a policy-rate path. What we will say is what it costs: the article is our only source for the Goldman view, and a source that carries two incompatible numbers for a cut that has already happened is a source whose arithmetic we cannot lean on for the cut that has not. The Goldman forecast is therefore reported here as one reader’s account of one house’s view, and nothing stronger.
This is the second week-ahead number in two days to collapse on contact
Yesterday morning the same thing happened to Bank Indonesia. A 75 basis point hike had sat in our notes for five runs, sourced to the same 16 October week-ahead. A second reader appeared 45 minutes before the pre-London slot fired and said 25 basis points, to 6.00 per cent — 50 basis points apart on the size and 50 on the destination, on a decision inside 24 hours. Today a 175 basis point cut turns out to be 75 basis points from the nearest consensus and 175 from two houses.
Two instances is not a rate, but the mechanism is identical and it is worth naming: a week-ahead entry looks like a schedule and behaves like a note. It carries a central bank, a direction and a size, which is the shape of a calendar row — and it carries no release time, no timezone, no forecaster, and no indication whether the size is a consensus, one house’s call, or an editor’s recollection. Ours did not survive either test it was finally put to.
The operational version, for anyone keying a system off a week-ahead: a size with no attributed forecaster is not a consensus, and the absence of a second reader after six askings is itself the finding. We should have dropped this on the third asking rather than the sixth.
We cannot confirm the time either, and it came from the same entry
Our notes have the decision at 11:00 UTC on 22 October. None of the three readers we opened today gives a time. One names Thursday without a date. One names 22 October without an hour. One references the meeting without either.
So the 11:00 timestamp rests on exactly the entry whose headline number we have just dropped. If you have a news window scheduled around 11:00 UTC on Thursday, it is sitting on single-sourced metadata from a source that has now been wrong about the size by 75 to 175 basis points. Widen it, or go to the issuing institution before the decision. We did not, and that is a gap in this article rather than a fact about the world.
What this does not tell you
It does not tell you what the central bank will do. Three readers say 100 and two houses say zero, and this desk has spent three weeks demonstrating that a settled consensus dispute still is not the print. If you want a view with a track record behind it, prefer a conditional threshold to anyone’s point estimate.
No primary source was read. The central bank’s own announcement page, its decision history and its Survey of Market Participants were not reached — the survey figures above are a secondary reader’s account of them, and the 37.00 per cent policy rate, while given by three readers, is not confirmed at the issuer. Every figure in this article is a secondary reading.
The decision time is unconfirmed, as above. We have no lira level: one reader says the currency slid to new lows and gives no number, and we did not price it independently, so we cannot tell you what is already in the exchange rate. We also cannot tell you where the 175 basis point figure came from originally — only that it is in our notes from 16 October and that nothing we opened today supports it.
And the two houses expecting a hold are two readers’ accounts of two houses, published 9 and 19 October. Neither has been restated since, and one of them carries two incompatible figures for a cut that has already happened.
