Dollar-yen opens the Tokyo week at about 159.70, which is thirty pips under a strike that comes off the board today. The option expiry board we have carried since 11 November put New York cuts on the 160.00 level on the 16th and the 17th — today and tomorrow — and in the three sessions since, spot has been through 160.00 twice and finished Friday below it both times. Friday ran 160.19 in Tokyo and 159.52 in New York, a sixty-seven-pip peak-to-trough against a one-week implied volatility that prices seventy-one pips for a whole day, and it closed in a six-pip band two publishers disagree about. The temptation is to read all of that as a yen story with an intervention warning attached. We think the positioning evidence says it is a dollar story that the yen happens to be the loudest expression of, and that distinction decides whether you size this week off a Tokyo headline or off a Treasury yield.
The strike you are trading into expires today, and the board that would tell you its size does not exist yet
Three facts, in the order they matter. Spot opened the Monday session at about 159.70 on the first Tokyo read we could find, stamped 06:05 Tokyo time — 21:05 UTC Sunday. The overnight high was about 160.06. And the 160.00 New York cut, which our 11 November board placed on the 16th and the 17th, is today.
- Monday open 159.70 against the 160.00 strike: thirty pips, or 0.42 of a one-day deviation. At the 7.09 per cent one-week implied volatility we read on Friday, one standard day on this pair is 71.3 pips. Thirty pips is not a move; it is a drift.
- Friday’s Tokyo high of 160.19 is forty-nine pips above the open, 0.69 of a deviation. The 2 September high of 160.39 is sixty-nine pips above, 0.97 of a deviation — one ordinary day, exactly.
- Friday’s peak-to-trough was sixty-seven pips, 0.94 of a standard day, delivered as a push through 160 in Tokyo and a slide to 159.52 in the New York afternoon.
What we cannot give you is the notional sitting on that strike, and for the tenth consecutive session the reason is arithmetic rather than sourcing. The only publisher we have found that posts strikes with size bands posts at about 12:40 UTC on a trading day. This run read the board at 23:09 UTC on a Sunday. Monday’s board will not exist for another 811 minutes, and the cut itself is 951 minutes away. Three routes were tried and all three correctly returned nothing, because nothing had been written. We have spent nine sessions calling this a fetch problem; it is a schedule problem, and we published the subtraction on Friday rather than trying a fourth route.
So treat the strike as a date, not as a size. A cut date with an unknown notional still tells you something a price chart does not: there is a known reason for someone to care about 160.00 between now and the New York afternoon, and it stops caring tomorrow.
Two official warnings, and price went the other way — because the move is not in the yen
The Ministry of Finance’s Vice Minister for International Affairs warned on 12 November. The Finance Minister warned on 13 November, saying moves had been one-sided and rapid and that Japan would stay in close contact with the United States Treasury. Between those two dates spot cleared 160.00 for the first time in about two and a half months and then did it again, higher.
The cross-sectional evidence says the yen is not the active leg:
- Euro-dollar fell for a fourth consecutive session to close at 1.1023, with a low of 1.1020 described by two Tokyo readers as the weakest in a year and seven months. On a 6.15 per cent one-week volatility, euro-dollar’s standard day is 42.7 pips, so a fourth consecutive fall is a trend in the quiet pair, not a shock in the loud one.
- The dollar index is at or near its own year-and-seven-month high on the same reading. That is a characterisation rather than a level, and we flag it as one: no publisher we read on Monday morning gave the index a number.
- The ten-year Treasury yield is either 5.10 per cent or 5.17 per cent depending on what you are asking for, and that is not a vendor dispute. The 5.10 figure is a settled daily series stamped 13 November 19:00 New York time — 14 November 00:00 UTC — carrying a twelve-basis-point one-day rise. The 5.17 is a Tokyo desk’s Monday-morning characterisation of roughly a one-month high. A close and an intraday extreme are different objects; the seven-basis-point gap between them is the difference in the question, not an error in the answer.
If the dollar is the active leg, then an intervention sized to defend a yen level is being asked to stand against a repricing that is happening in every dollar pair at once. That is a materially worse trade for the authorities than the same level reached on a yen-specific move, and it is why we would not read the warnings as a timing signal. One Tokyo desk made the stronger version of this argument on Monday morning — that the warnings simply did not stop the rally because they were aimed at the wrong leg — and we are relaying that as one house’s view, because it is a causal claim and we have one reader for it.
The repricing that actually moved is one meeting further out, and we have one source for it
December is described as fully discounted on the one feed anybody names, which is where it has been for days. The number that moved over the week is the meeting after it. One Tokyo research note on Monday morning puts the probability of a further hike at the February meeting at over seventy-one per cent, against just under sixteen per cent a week earlier, naming CME FedWatch as the feed.
That is a rise of at least fifty-five points, a factor of at least four and a half, in a week. We are publishing it with two labels attached and we want both read. It is single-sourced, and it reaches us second-hand through a publisher quoting a feed. Our own file holds a December half-point tail at 10.40 per cent on 12 November and 14.5 per cent on 13 November, both from that same feed by different routes, and we have no independent February reading at all to check this against.
Why it matters more than the December number for anyone running an option book: a hike in December is inside a one-month option and a hike in February is not. Our Friday surface showed three of seven pairs pricing one-week volatility below one-month, and the mechanism we gave was the December meeting sitting inside the longer tenor. If the market has spent a week moving a February probability from sixteen to seventy-one, that mechanism is migrating out the curve, and a term structure read on Friday’s logic will be wrong about which tenor carries the event. We have not re-read the surface this morning and so we are not claiming it has moved. We are claiming the reason to re-read it has changed.
What a 130-pip forecast range is actually telling you
One Tokyo house published a Monday range of 159.000 to 160.300. Measured against the volatility surface that is a width of 130 pips, or 1.82 one-day deviations, with the open sitting seventy pips off the floor and sixty pips off the ceiling. The 160.00 strike sits thirty pips inside the ceiling.
A range that wide is not a forecast, it is a statement that the publisher expects an ordinary-to-slightly-busy day and will not pick a direction — which, with a GDP print forty minutes away and a hawkish board member speaking in the afternoon, is a defensible thing to decline. The useful part is the asymmetry it implies. If the ceiling is 160.300 and the strike is 160.00, then the band the publisher has drawn puts the expiry level inside the plausible range rather than at its edge. That is the shape under which an expiry actually matters: close enough to be reachable, far enough that reaching it is a decision rather than noise.
The uncomfortable version, and we would rather say it than not: if you are sizing a position this morning off the fact that 160.00 has been rejected twice, you are using two data points to describe a level whose only mechanical support — the option cuts — stops existing tomorrow night.
What this does not tell you
- We do not have the notional on the 160.00 strike, and we will not have it from any run on the current schedule, because the only publisher that carries sizes posts 811 minutes after this read and its sizes sit behind a paid membership even then. Tenth consecutive session. The cut dates we are using come from an 11 November board, five days old.
- The February probability is one reader, second-hand. We have no independent confirmation of either the seventy-one or the sixteen, and the week-ago figure in particular is that same publisher’s own prior. If it is wrong, the whole third section of this piece is wrong.
- Friday’s close is two different numbers — 159.72 at one house and 159.66 at another, six pips apart. We have used 159.70 as the Monday open because that is the figure a Tokyo reader stamped at 06:05 local, and we have not resolved the close.
- Friday’s Tokyo high is 160.18 or 160.19, one pip apart across two readers. We published 160.19 on Friday from two publishers and a third now gives 160.18. One pip is not a dispute and we are not adjudicating it, but the record should show the figure moved.
- No publisher we read on Monday morning gave the dollar index a number, only a characterisation. That is the second consecutive session in which the index has driven our argument and we have printed no level for it.
- We have not re-read the volatility surface or the risk reversal this morning. Every volatility figure in this article is Friday’s 07:30 UTC reading, and the one-week tenor on this pair moved 1.20 points in two days last week. Treat the 71.3-pip standard day as five sessions stale at the front end.
- The claim that the dollar rather than the yen is the active leg is an inference from three cross-rates, not a measurement. We have not computed a correlation and we are not going to pretend three levels are one.
