This morning this desk published a 50 basis point disagreement with a deadline attached: two readers had Bank Indonesia going 25 basis points to 6.00 per cent or 75 basis points, both of them up, resolving at about 07:20 UTC. The decision has landed and both readers are wrong on the headline rate, which was held at 5.75 per cent for a fourth consecutive meeting. The 25 basis points was real. It arrived on the deposit facility rate, which went from 4.75 to 5.00 per cent, and neither of our readers mentioned that instrument existed. The question we asked was how big. The question that decided the answer was which rate.
What was decided, against what we had on the board
The central bank held its seven-day reverse repo rate at 5.75 per cent, raised the deposit facility rate by 25 basis points to 5.00 per cent, and left the lending facility rate unchanged at 6.50 per cent. Two readers give the same three figures.
Our published position, carried from 16 and 20 October, was a 75 basis point move from one reader and a 25 basis point move to 6.00 per cent from another, with the direction two-sourced as an increase on a sliding currency. Score it honestly. On the headline rate the nearer reading missed by 25 basis points and the farther one by 75. On direction, both were right that policy tightened and both were wrong about where. And the reader that named 25 basis points named the correct size on the wrong line of the same press release.
That is a more useful failure than a simple miss, because it is not a forecasting error at all. Both readers forecast the policy rate, which is the field every aggregator carries and the only field most news filters parse. The move happened in the field nobody quotes.
The corridor went symmetric, and that is the event
Run the arithmetic on the three levels rather than the one. Before today the corridor around the policy rate was 100 basis points wide below and 75 above: deposit facility at 4.75, policy at 5.75, lending at 6.50. After today it is 75 and 75. The corridor is symmetric for the first time in this sequence, and it got there by lifting the floor rather than the middle.
The deposit facility rate is where surplus overnight cash is placed with the central bank. Raising it by 25 basis points raises the effective floor under domestic overnight money by 25 basis points, whatever the headline does. So the honest one-line summary of today is not that Indonesia held. It is that Indonesian overnight rates have a floor 25 basis points higher than yesterday and a policy rate in the same place.
For anyone running a schedule-driven system this is the transferable part. A filter keyed on the policy-rate field records a non-event and moves on. A filter keyed on the direction of short-term funding records a tightening. Both are reading the same release correctly, and they disagree because the release has three rates in it and the calendar entry has one.
The consensus was a 52-to-48 split, which is not a consensus
The wire reports a Reuters poll in which 15 of 29 economists expected the hold and the remaining 14 expected a hike of at least 25 basis points. That is 51.72 per cent against 48.28 per cent.
So the majority view was correct on the headline rate by two votes out of 29, and our two readers were both in the 48.28 per cent. This desk has spent three weeks arguing that a consensus is not one number. Today adds the sharper version: a consensus can be a near-tie, and a vendor field reporting it as a single expected value destroys exactly the information you needed. A 52-48 split and a 90-10 split both print as one forecast, and only one of them is a forecast you should size against.
We also note what the poll was not asked. Nothing in what we read gives any count of economists expecting a deposit facility move. The split that resolved this release was never polled.
Why the context is a currency and an index provider, not inflation
The stated rationale in both readings is currency stability and capital inflows, reinforced by further incentives for hedging instruments under the domestic non-deliverable forward scheme, with the inflation target given as 2.5 per cent plus or minus one point for 2026 and 2027. One reader records that the central bank raised rates by 100 basis points across three moves between May and June to defend the currency, and that the rupiah has been hovering near an all-time low.
The acute trigger is not a macro release. One reader reports that an index provider warned last week it could relegate Indonesian equities to standalone status, which set off a selloff in both the currency and the stock market, with a record two-day rout in equities. A reclassification threat is a flow event with a date attached, and it is the kind of thing a macro calendar does not carry at all.
Which is the Asia-hours point. A regional central bank restructured its policy corridor during the Tokyo session, in response to an index-provider notice rather than to data, and dollar–yen traded 158.035 to 158.392 — 35.7 pips, on two publishers’ readings that agree to a tenth of a pip. The yen did not notice, and there is no reason it should have. But a session that quiet is the session in which a correlated Asian risk event is most likely to be mispriced by a system sized on yen volatility alone.
What this does not tell you
Bank Indonesia’s own release page was not reached in this run. Every figure above — all three rates, the poll counts, the inflation target and the hedging-scheme reference — is a secondary reading of the central bank’s announcement, taken from two independent publishers that agree on the three levels. The prior deposit facility rate of 4.75 per cent is an inference from a stated 25 basis point increase to 5.00, not a figure either reader printed on its own.
We cannot give you an announcement timestamp. The 07:20 UTC hour we published this morning was never confirmed at the issuer; one wire’s file stamp is consistent with something shortly before 07:40 UTC, and the second reader we used published at 10:19 UTC, three hours later. Treat the 07:20 as an estimate that has still not been settled.
We are not naming who announced it. One of our two readers calls the speaker Governor in its body and Deputy Governor in its own photo caption, and with no primary source read we will not pick between a publisher and its own caption. We also have no rupiah level: one reader describes the currency as near an all-time low and neither prints a rate in its body, so the number you want is not in this article.
Finally, the corridor argument is about mechanics and not about effect. A 25 basis point floor under overnight cash is a change in the price of domestic liquidity; whether it does anything for the currency is a separate question that today’s release cannot answer and this desk has not tested.
Related
- FX event calendar — the week’s remaining central-bank decisions in UTC, including Thursday’s Turkish decision at an hour we still cannot confirm.
- Signals — what a near-tied consensus does to a release-reaction rule.
- EA presets — Asia-session exposure settings when a regional policy event lands in thin conditions.
- 1KPIPS News — this morning’s piece that published the 50 basis point disagreement and its deadline.
