The Bank of Japan’s own release calendar, updated 23 October, gives tomorrow’s policy decision a date and then, where the time should be, the word “not yet set.” The Governor’s press conference has a clock — 15:30 Japan time, 06:30 UTC. The decision that actually moves the yen does not. And several hours before it, at 23:30 UTC tonight, Tokyo prints October consumer prices. Your filter has one event with a centre and one without, in the same currency, on the same morning, in that order.
The One Event in the Yen’s Own Currency Has No Clock
The issuer’s page is unambiguous about what it does not know. The Monetary Policy Meeting runs 29–30 October. For 30 October it lists the decision itself with the time marked undetermined, and the Outlook Report’s basic view the same way. The two items on that page that do carry times are the ones that do not set policy: September payment trends at 11:00 Japan time, and the Governor’s press conference at 15:30, live stream planned. The full Outlook text follows on 2 November at 14:00.
This desk has spent six weeks cataloguing broken calendar fields — a consensus in the wrong field, a rate without its instrument, an average without its window, a clock without its zone, an event without its type, an event without its time. This is the second instance of that last one, and it is the honest version. The Bank of Japan is not publishing a false precision for us to convert; it is telling us the field is empty. Our own notes already say an issuer that writes “not yet set” is more useful than one that writes a minute with no zone, and this is why: the first instructs you to widen, the second lets you believe you are narrow.
So the correct pad around tomorrow’s decision is not a symmetric window around a guess. It is a block that opens when the Tokyo lunch break would normally end and closes on an artefact — the statement hitting the wire — rather than on a clock. If your scheduler cannot express “block until event observed,” the honest fallback is to stand down from roughly 02:00 UTC to the press conference at 06:30 and pay for the dead hours.
Tokyo CPI Lands First, and the Ordering Is the Part You Can Trade
Tokyo’s October consumer price index publishes at 08:30 Japan time on Friday, which is 23:30 UTC tonight. The cadence is not in dispute: the September Tokyo print was released on 2 October at 08:30 Japan time, and two separate preview and review pages from the same publisher carry that date and hour.
That puts a leading national inflation indicator in the book several hours before a central bank decides, on the same morning, with the decision’s time unpublished. The asymmetry is the whole point. The CPI has a centre you can pad around to the minute. The decision does not. A news filter built on symmetric windows will handle the first correctly and the second by accident, and the gap between them is exactly the interval in which a Tokyo desk has new inflation data and no policy response yet.
We are not going to tell you the CPI will move the decision. The Committee’s votes are cast before the number is public. What it moves is the press conference, and the press conference is the item on that calendar with a clock on it.
Last Night’s Tape: the High Came Before the Fed, Not After
Dollar–yen made its high at 158.35 — one publisher prints it to three decimals, 158.344 — and that high was on the board before the Federal Reserve announced. A second reader at 03:31 Japan time, thirty-one minutes after the decision, had the pair at 158.30 and wrote that the currency market showed no major reaction. By 06:51 Japan time the first publisher had it pushed back to around 157.77, and the Tokyo session opens in the high 157s.
Both readers were right about what was in front of them. The reaction did not arrive in the thirty-one minutes after the statement; it arrived across the press conference and after it. Roughly fifty-eight pips of give-back, read from two timestamps rather than a tick history, which is why we are quoting the two timestamps rather than a single “daily change.”
The level that mattered on the way up was not an option strike. The 200-day moving average sits at 158.54 on the same publisher’s read, and the high stopped nineteen pips underneath it. We carried a 1.1 billion dollar strike at 158.00 into yesterday’s New York cut, and the session subsequently traded through it to 158.35 — but the cut was at 14:00 UTC and we cannot read the print at that minute, so that is a strike we watched trade through hours later, not a pin or a break we scored. We are saying so rather than claiming the result.
Eighty-Three Per Cent Is Not a Probability
A Bloomberg survey released on 21 October puts 83 per cent of economists on a Bank of Japan hike in December, up from 35 per cent a month earlier. The policy rate is 1.25 per cent and the survey’s median terminal forecast is 1.75. Named in it: Keiji Kanda at Daiwa Institute of Research, looking for December and March; Takeshi Yamaguchi at Morgan Stanley MUFG, who calls this month a hawkish hold; Naomi Muguruma at Mitsubishi UFJ Morgan Stanley.
This desk has carried, single-sourced and for three runs, a figure of 75 per cent for a 25 basis point move to 1.50 on 18 December. It is tempting to treat 83 as the second reader that finally closes that item. It is not, and the reason is the kind of thing this desk keeps finding in its own files.
Eighty-three per cent of economists expecting a hike counts opinions. Seventy-five per cent for a hike prices an event. They are different objects with the same units, and a system that stores both in a field called “December hike probability” has just averaged a vote with a price. A forecaster survey can read 100 per cent while the market prices 60, and neither is wrong. So the open item stays open: we still have one probability reading for the December Bank of Japan move, and we now also have a survey, which is a different measurement of a different thing.
What This Does Not Tell You
It does not tell you where dollar–yen trades this session. Our levels come from two publishers at two timestamps, and we have no tick data and no verified close. It does not tell you what the Bank of Japan will do: a hold is the broad expectation and we have not sourced a probability for tomorrow specifically, only for December. It does not tell you the decision time — that is the article’s subject, and we cannot fill the field either. The 158.344 high, the 157.77 pullback and the 158.54 moving average are each one publisher’s reading. We did not reach the Tokyo Metropolitan Government’s own price statistics page for the CPI cadence and relied on a broker’s dated preview and review pages instead.
