Japan Exchange Group’s own calendar lists today, Tuesday 3 November, as a non-trading day: “Nov. 3 (Tue.) Culture Day.” No Tokyo cash equity market, no cash JGB market, and Japanese banks closed with it. Into that hole the Reserve Bank of Australia delivers a decision at 03:30 UTC and a press conference at 04:30, and the first American polling stations do not close until 23:00 UTC tonight. So the one central bank deciding anything today decides in the hours Tokyo normally supplies the depth, on a day Tokyo is not there, and the whole session runs before a single midterm result exists. This is a size question, not a direction question, and the answer is smaller.
The exchange is shut, and that is the issuer saying so
We went to Japan Exchange Group rather than an aggregator. Its English trading calendar, last updated in February, states that “JPX markets are closed on Saturdays, Sundays, national holidays, and on the dates indicated below,” and the row reads “Nov. 3 (Tue.) Culture Day.” Two secondary calendars describe the same day as a full closure with no shortened session. The next two after it are 23 November, Labor Thanksgiving Day, and 31 December.
What follows from a national holiday is that Japanese banks are not open, and two secondary sources say so in those words, one of them adding that many Japanese banks treat national holidays as non-business days for interbank settlement. From that we infer there is no 09:55 Tokyo fixing today. We want to be exact about the status of that: it is our inference from bank closure, not a statement any source made. No page we read mentioned the fixing at all.
It matters because the fixing is the one scheduled flow that reliably gives the Tokyo morning a shape. Strip it out and the session before an Australian rate decision is a tape with no domestic anchor, no local cash market to hedge into, and no Japanese official sitting at a desk to comment on the currency. Dollar–yen will have a price all day. It will not have the usual reasons for that price.
One decision, inside this slot’s own window
The Reserve Bank announces at 03:30 UTC — 14:30 Sydney, which is AEDT and UTC+11 today — with the Governor’s media conference an hour later at 04:30, and the quarterly Statement on Monetary Policy released at the same meeting. Two independent readers give those times identically, and a third source we read yesterday put the decision at 12:30 Japan time, which is the same 03:30 UTC. The time field is as settled as anything in this file gets.
The cash rate is 4.60 per cent, last raised by 25 basis points on 29 September in a decision both readers describe as unanimous. A hike to 4.85 per cent is the expected outcome, which would be an 18-year high by one reader’s description, and the same reader notes the September press conference was more dovish than usual — so the statement’s language, not the number, is the part with information in it.
For a system this is a window problem with a clean answer. 03:30 is the decision and 04:30 is the press conference, both inside the Asian session, both on a day when the second-largest currency in the main crosses has no home market. If you run a news filter, it needs two entries today and not one, and the second is the one with the Governor in it.
Four feeds, four numbers, one binary
We read four readings of the same event and got four answers. One wire at 07:15 UTC yesterday had money markets at “around an 81 per cent likelihood” of a 25 basis point hike and around 19 per cent for a hold. A rate-watch page timestamped 07:02 Sydney, which is 20:02 UTC yesterday, had ASX 30-day futures at 83 per cent hike, 17 per cent hold, zero cut, and separately gave money-weighted prediction markets at 88 per cent hike, 11 per cent hold, 1 per cent cut. A bank’s Asia note said about 80 per cent priced per swaps.
Eighty to eighty-eight, across futures, swaps and prediction markets, at three different hours. We are not going to publish one of them as the market-implied probability, because this channel has made that mistake before and the honest statement is that the band is the reader’s error bar. Note also that the same rate-watch page implies a post-meeting cash rate of 4.81 per cent from futures, which is not 4.85 and is not supposed to be — it is a futures-implied average across the month rather than a target level. Two numbers, two units, and the same page carries both.
The repricing is the part we would actually trade off. The same futures series read 20 per cent for a hike on 2 October, 62 per cent on 26 October and 83 per cent now. That is not a market that has believed this for a while. It is a market that changed its mind inside five weeks, and the December meeting on 7–8 December is already being quoted at an expected 4.93 per cent by the same page, which means a hike today is not treated as the end of anything.
The yen position is crowded, and the measurement is six days old
This desk has owed a piece on positioning measurement for weeks and has deferred it repeatedly. Here is the number with its date attached. A Tokyo research note published yesterday morning reads the CFTC’s yen futures as showing speculators’ net long at 101,591 contracts as of 27 October, up from a smaller figure earlier in the month, under a chart captioned as Chicago currency futures non-commercial positioning.
Three things have to be said in the same breath as that figure. It is measured at the Tuesday close, so as of today it is six days old and was already stale when it was published. It is one reader of the CFTC’s release, not the release itself. And net long yen means short dollar–yen, which is the position that has been wrong for most of the last fortnight: the same note has the pair at a 158.344 high last week against a 156.651 low, and dollar–yen was 157.88 in the New York afternoon yesterday by one reader and 157.73 in an earlier Monday snapshot by another.
A crowded short in the pair whose domestic market is closed, on a day with an offshore central bank decision inside the thin hours, is a specific and uncomfortable combination. It is not a reason to be long. It is a reason to assume that whatever moves today moves further than the news in it justifies, because the people who would normally absorb it are not at work.
Where the levels are, and the two we will not quote
Dollar–yen sits between its averages. One reader gives the 25-day at 157.881 and the 200-day at 158.573 as of yesterday morning, with a 14-day RSI of 50.66 — the middle of the range on both counts — a weekly pivot at 157.509, and a forecast range for the week of 155.80 to 159.20 with a central band of 156.65 to 158.35. October’s own high and low were 158.949 and 156.078. A separate note from the same publisher yesterday evening gives a November range of 155.000 to 160.000 and describes the trend as searching for direction, which is an unusually candid thing for a forecast to say and we are quoting it rather than tidying it.
Two figures we are refusing to state as single numbers. The Australian dollar: 0.7029 in an early Monday snapshot and 0.6998 in the New York afternoon, 31 pips apart. Those are two hours, not two opinions, but we are going into an Australian rate decision and we are not going to print one of them as the level the market is at. Crude: $102.31 and down 1.7 per cent in that same early snapshot, then $96.12 and down 7.68 per cent in the afternoon. Read in sequence that is one session that kept falling, not a vendor dispute — but the close is single-sourced, so we publish the direction and the band and not a settlement.
A near-eight per cent fall in crude matters to this pair specifically, because Japan imports its energy and a cheaper barrel is a yen-positive input. It arrived yesterday. Nobody in Tokyo can act on it today. Also carried from yesterday morning, and single-sourced from a headline whose body is members-only: sell orders reported at 158.00 in dollar–yen with stop-loss buying above. That is 12 pips above the afternoon reading and it is the nearest thing to a structural level anyone has published for today.
What this does not tell you
We have no Tokyo opening print, because this article is written before the session and because there is no exchange opening today. We have no close for yesterday — both FX readings above are afternoon snapshots whose own publishers say they are not closes, and the attempts we made at two vendors’ four-value tables for 2 November both failed to load. So nothing here is anchored to a settlement.
No cross rate is sourced. We have no Australian dollar–yen, euro–yen or sterling–yen level for yesterday afternoon at all, which is awkward given that the argument above is about crosses, and we are not deriving them from the two dollar legs because the legs are timed differently. Australian dollar–yen is the pair where both sides of this setup meet and we cannot tell you where it is.
The rate-hike pricing comparison we would most like to make is single-sourced and we are labelling it rather than leaning on it: one Tokyo note reads Bloomberg’s overnight index swaps at around 79.8 per cent for a Bank of Japan 25 basis point hike in December against 74.8 per cent for the Federal Reserve, as of mid-morning Tokyo yesterday, having been near 63 per cent for the Bank of Japan on 23 October. One reader, one feed, and a comparison of two central banks that deserves two readers before anyone sizes off it.
And the midterms are not analysed here. Poll closings run from 23:00 UTC tonight through about 06:00 UTC tomorrow on the timetable we read, first results land in tomorrow’s Asian and European hours, and the pre-vote commentary we looked at was dated 21 and 29 October with prices to match. The only thing we are claiming about the election is a timing fact: none of it resolves during today’s Tokyo session, so today is the last thin session before an unpriced binary rather than a session that contains one.
