Dollar–yen traded 158.65 down to 157.83 across three sessions on 13 October — eighty-two pips, with a fifty-four-pip reversal inside it on a wire report about the Bank of Japan — and then handed almost all of it back. New York closed at 158.33, fifty pips above the London low and thirty-two below the Tokyo high. At 08:00 JST this morning the Fisco desk put the pair at 158.34 with a session range of 158.32 to 158.39. Seven pips. The day before had moved eighty-two. Tokyo has opened eleven and a half times quieter than the day it is reacting to, twenty-two pips below the 200-day average, and it is doing so on purpose.
Eighty-two pips in three sessions, and the close landed in the middle of them
Take the sequence as the three sessions actually produced it. Tokyo ran 158.65 to 158.19 on 13 October, a forty-six-pip session that handed back Monday’s holiday break. London took it lower: broad dollar selling pulled the pair toward 158.00, the Bank of Japan report lifted it to about 158.37, and then the selling resumed and took it through to 157.83. New York bought it back into the 158.40s and closed at 158.33.
The arithmetic of that close is the whole point. It sits fifty pips above the session low, thirty-two below the session high, and twenty-eight below Monday’s close. Kyodo’s 07:00 JST quote this morning — 158.28 to 158.41 — was seven sen below its own reading of twenty-four hours earlier. An eighty-two-pip range delivered a seven-sen day-on-day change at a fixed hour, and a twenty-eight-pip change close to close.
If you trade this pair on a daily candle you have a small red bar with long wicks. If you traded it on a fifteen-minute chart you had two moves worth having and one of them was in the wrong direction first. Those are not two readings of one market. They are two different markets, and which one you were in was decided by your holding period rather than by your view.
The report that moved it fifty-four pips had already been published as consensus
The move’s stated cause was a report, carried by two publishers at 08:49 and 09:22 UTC and tracing to one originating wire, that Bank of Japan officials are inclined against a consecutive rate increase at the 30 October meeting, with December still live. We said yesterday that this is two readers of one source and not two sources, and that the Bank itself has said nothing. That still holds.
What is new this morning is the timestamp. Gaitame.com’s weekly outlook was published at 15:53 JST on 13 October — 06:53 UTC, one hour and fifty-six minutes before the first reader of the report — and it already described an October increase as almost fully priced out and put December at over eighty per cent. Our own file carried five mutually incompatible October probabilities, between roughly ten and twenty-five per cent, before any of this.
So the market was handed, as news, a thing a retail broker had published as settled consensus two hours earlier. And it produced thirty-seven pips of yen weakness and then fifty-four pips of yen strength through the starting point. That is not a market learning something. That is a market with a thin book being asked to re-price a confirmation, overshooting, and then being leaned on by whoever was actually trying to buy yen. Treat the first thirty-seven pips of any Asia-hours policy headline as the least informative part of the move. It is the part that happens before anyone with size has decided anything.
The pair traded both sides of the 200-day in one day, and the publisher moved its own number by a pip
This desk has been arguing about the dollar–yen 200-day for a week, and 13 October gave the argument a clean answer and a new embarrassment at the same time.
The clean answer: the day traded both sides of it. Against the simple readings we now hold — 158.50, 158.52, 158.53, 158.54, 158.55 and 158.56, a band six pips wide — the 158.65 high was nine to fifteen pips above and the 158.33 close is seventeen to twenty-three pips below. There is no convention in our file on which that day was one-sided. Against our own exponential reading just under 158.00, now six days old and never recomputed, the close is thirty-four pips above and nothing at all happened.
The embarrassment belongs to the measurement, not to the market. Gaitame.com published the 200-day at 158.56 twice on 13 October, at 07:34 and again in the 15:53 weekly outlook. Its own technical page at 02:15 JST this morning gives 158.55, and gives the previous business day’s value as 158.55 too. One publisher, two surfaces, one pip apart, and the back-value disagrees with what the same publisher printed in prose on the day in question. Neither page names the convention. The running tally for the specification we published on 8 October — name the convention, the publisher and the time — is now one labelled reading against five unlabelled ones.
The same page is worth reading for what it does to the rest of the picture. The five-day average is 158.25, eight pips below the close; the twenty-one-day is 157.56, seventy-seven below; the ninety-day is 159.46, a hundred and thirteen above. The twenty-day Bollinger band at two standard deviations spans 156.24 to 159.13 — two hundred and eighty-nine pips — and the close sits seventy-two per cent of the way up it. Every short average says the pair is above trend, every long one says it is below, and the single level that sits closest to spot is the one nobody will label.
A seven-pip box is a session declining to price an American number
Tokyo is not quiet because nothing happened. Tokyo is quiet because the thing that matters happens at 12:30 UTC tonight, in New York, and it is American.
US September consumer prices print then. The consensus this desk established before the event runs 3.6 or 3.7 per cent on the headline against a 3.4 prior, and core at 2.39, 2.4 or 2.5 against 2.4, with every reader we have putting core at 0.2 per cent on the month. Gaitame’s weekly, which names no source for it, gives headline 3.6 to 3.7 and core around 2.5 — consistent with the top of our range on core and with our headline. Before that, Japan’s September M2 money stock at 23:50 UTC, which the Fisco desk itself expects to do nothing.
The useful way to hold a seven-pip box is as an inventory statement rather than a view. Nobody is taking risk into the print, which means the book into 12:30 is thin on both sides, which means the first move will be larger than the information in it — exactly as the first thirty-seven pips of yesterday’s BOJ headline were. If your system sizes off realised range, a seven-pip Asian session will tell it this is the calmest market of the week and it will size up accordingly. That is the error. The realised range is low because the event has not happened, not because the event is small.
Two levels frame it without needing a direction. Above, the 200-day band at 158.50 to 158.56 and then the 158.75 high of 12 October, which Gaitame explicitly nominates as the upper dividing line. Below, 157.83 from yesterday and then last week’s 157.43 low. The pair is currently sitting almost exactly in the middle of a hundred-and-thirty-pip corridor, which is the one honest description of it.
What this does not tell you
It does not tell you which way the pair breaks, and nothing in this piece is a directional call. It does not tell you that the Bank of Japan will skip October: the Bank has said nothing, the report has one originating source behind two readers, and the 30 October meeting is sixteen days away.
It does not give you a settled close for 13 October. We have 158.33 from one publisher and a late-session 158.30 from another, three pips apart, and no venue’s official fix. Monday’s 158.61 close, from which the twenty-eight-pip change is computed, is carried from our own file rather than re-read today. Our exponential 200-day reading is six days old and this desk has still never computed a 200-day average from a price series of its own, which means every simple reading quoted here is somebody else’s arithmetic.
It does not establish a cause for the overnight recovery from 157.83. Two Fed officials spoke into the New York session — the Boston president backing another increase before year-end and the New York president saying one more this year is appropriate with none needed in 2027 — and crude rose, but we read two publishers on crude and they are a dollar fifty-six apart on the level and more than a point apart on the percentage change, so we are not naming a cause we cannot price. We looked.
And no option expiry ladder is available for tonight’s 14:00 UTC New York cut. The index that carries them returned a client error for the third consecutive slot. Treat 14:00 as a window, and assume the unknown there is large rather than zero.
Related
- FX events calendar — the 12:30 UTC print and the 14:00 cut in one view.
- Signals — how a range this narrow should change position size rather than direction.
- EA presets — news-window and realised-range filters for an Asian session sitting in front of an American print.
