Twenty-four hours ago this desk published that the only publisher with an option board had tagged dollar–yen 158.00 to a New York cut on both the 5th and the 6th, called it “extremely large”, and we wrote a section arguing that today’s board was the bigger of the two. This morning the same publisher put out three order items and 158.00 is not among them. The new dollar–yen level is 157.80, tagged to a cut on the 12th. And the Tokyo session high printed 158.00 — the strike nobody republished, to the figure, on the one publisher’s reading that gives a range at all.
The tag reading is now settled and today’s board is gone
Yesterday we flagged one inference explicitly as an inference: that the expiry tags on these items are days of the month. Today closes it. The three items, from Traders Web FX’s public index, timestamped in UTC:
- 21:22, 5 November — dollar–yen order at 157.80, “OP 12th NY cut”. No size qualifier.
- 21:33, 5 November — euro–dollar order at 1.1215, option, 12th New York cut. No size qualifier.
- 21:41, 5 November — euro–yen 177.10, stop-loss selling on a break below. No expiry tag.
“The 12th” can only be a date. There is no reading of that number as a size, a strike or a contract code that survives contact with the other two items. So the day-of-month reading we were careful to label as unconfirmed is now about as confirmed as a public headline can make it, and the thing it confirms is awkward: the publisher’s fresh options content this morning points six days forward, not at today.
That does not delete today’s 158.00 expiry. Yesterday’s item named cuts on the 5th and the 6th, and nothing published since contradicts it. What has happened is weaker and more useful to know: the board was published once, yesterday, and has not been reconfirmed, while the level itself has changed sides. At yesterday’s 158.46 New York close the strike sat 46 pips below spot. By 23:32 UTC the pair was around 157.32, which puts the same strike 68 pips above. An “extremely large” strike underneath the market and the same strike above it are different objects, and the only publisher who told us it was there has not mentioned it since.
The session high was the strike, and we are going to say why that is weak evidence
Tokyo traded a 75-pip range. One publisher gives it as 157.25 to 158.00 as of 03:18 UTC; another, at 02:48, describes a fall to about 157.26 and a recovery to about 158.00. The low differs by one pip between two readers and the high is the same number on both. The Tokyo fixing published by one bank at 01:06 was 157.77, three pips from this morning’s new 157.80 strike, and the 10:00 Tokyo quote an hour earlier was 157.75 to 157.77.
So the pair spent the session walking across the 157.80 level and stopped dead at 158.00 on the way up. If you want the tidy version, the board we were given yesterday described today’s range to the figure at the top and the publisher’s own morning order level sat inside it.
Now the uncomfortable part, because this is exactly the kind of coincidence a desk talks itself into. 158.00 is a round number. Round numbers are where resting orders cluster, where stops cluster, where option strikes get written, and where price stops for reasons that have nothing to do with any of the three. A recovery halting at the figure is overdetermined. We cannot distinguish an expiry pin from a plain round-number rejection with the material available, and we are not going to publish the inference that the strike did it. What is honestly established is narrower: the level was identified in advance, it was the session’s high, and anyone who had placed a stop just above 158.00 on the strength of a 200-day break would have been inside the one tick cluster the day produced.
And note what the pair did not do. Three Tokyo publishers have given the 200-day moving average as 158.60, 158.63 and 158.65 across this week. Today’s high was 60 to 65 pips below the lowest of them. One Tokyo desk at 23:36 UTC described the pair as having broken down through the 200-day with a heavy top as a result, and gave no value for the line. Against the Ichimoku reading our Asia desk published twelve hours ago — cloud floor 157.18, top 158.44 — the entire session fitted inside the band, seven pips above the floor at the low and 44 below the top at the high.
The trapdoor was already open when the sign went up
Yesterday we wrote that the euro–yen stop cluster at 177.10 sat ten pips below the floor of a published forecast range, and that “a forecast floor with a stop cluster underneath it is not a floor, it is a trapdoor with a sign on it”. Here is what happened to it.
Euro–yen fell to 176.83 in the New York session, 27 pips below the cluster, before this morning’s item was published. At 01:00 UTC it was quoted 176.99 to 177.01, still below. The Tokyo low on one reader’s range was 176.47, 63 pips through the level, and by 02:48 it had recovered to about 177.16, above it again. The call was right in direction and the honest scoring is that it was right before the item that described it was published: the 21:41 order note naming 177.10 as a break level arrived after the break.
That is a source-timing fault worth a rule rather than a complaint. An order page is a snapshot of resting interest at the moment somebody wrote it down, and the moment is not the timestamp. When the overnight session has already traversed the level in the item you are reading, the item is telling you where stops were, not where they are, and the two differ by exactly the flow that moved price. A preset that treats a published stop cluster as a live obstacle is sizing against liquidity that has already been consumed.
The one number that changed is the one you size with
Strip out the strikes and the single most consequential figure of the last twenty-four hours is a volatility quote. At 08:27 UTC yesterday one Japanese publisher had one-week dollar–yen implied volatility at 6.23 per cent and called it the lowest since 22 October. This desk published that figure and, twelve hours later, told you to treat it as stale because the Nasdaq had fallen 5.52 per cent in between. At 02:40 UTC today the same publisher’s volatility item reads, on its headline, that one-week dollar–yen is in the high seven per cent area.
Same publisher, same series, same tenor, two days. Call it roughly a point and a half, about a quarter more volatility priced into the week, and the week in question is the one containing the US October employment report at 13:30 UTC today. That is a cleaner comparison than anything else in this note because nothing about the measurement changed except the number.
The operational consequence is arithmetic, not opinion. If your position size, your stop distance or your option budget was derived from a 6.23 reading — which is to say from any reading taken before yesterday’s New York session — it is now calibrated roughly a quarter too tight for the week it has to survive. The strikes above tell you where not to put a stop. This tells you how big the thing you are stopping out of has become. Only one of those two is actionable before London opens, and it is not the board.
The board with the sizes exists, and it arrives after you need it
Three days of this desk hunting for notionals produced one finding today. Minkabu’s FX news index carries a daily item titled, in translation, “today’s New York option expiries — main strike prices and sizes”. Yesterday’s instance is timestamped 12:00 UTC. We have not read its body — the text did not come through on the fetch, and we do not know whether that is a paywall, a rendering fault or our own method — so we are reporting the existence of a titled item and nothing about its contents.
Take the schedule at face value and it is the wrong schedule. A board published at 12:00 UTC lands five hours after this note, two hours after the London open at 07:00, 90 minutes before payrolls at 13:30, and three hours before the cut at 15:00. If the sizes are in there, they are available in time to adjust an open position and never in time to decide whether to take one. The desk that reads the board at 12:00 has already traded the London morning off levels with no notionals attached, which is what we did yesterday and are doing again today.
That is a scheduling problem rather than a sourcing problem, and the two call for different fixes. We have been looking for a better publisher. What the file now says is that the best available publisher is on a clock that does not match the slot, and the honest response is to stop treating a pre-London expiry board as obtainable and start treating the 12:00 item as a position-management input.
What this does not tell you
We still have no notional, anywhere, for any strike. Everything above comes from public headlines on a members-only service. We cannot tell you whether the 158.00 expiry is a hundred million dollars or three billion, and we said yesterday that a strike without a size forbids a pin thesis. Nothing today changes that, including the fact that the session high was the strike.
Two of the figures here are headline reads. The high-seven volatility quote and the existence of a sized expiry board are both taken from item titles we could not open. The volatility comparison is only as good as the assumption that the publisher is quoting the same one-week series it quoted yesterday, which we believe and have not verified.
The order index did not move all session. Two fetches hours apart returned the same page with the same newest item, timestamped 23:52 UTC on 5 November. We cannot tell whether the publisher posted nothing after that or whether we were served a stored copy, and those are different faults with the same appearance. If it is the latter, there may be a 6 November expiry item we never saw.
We have no current level. Every price above is a Tokyo-session print, the freshest of them from 03:18 UTC. One desk’s 05:53 note says the recovery has been limited and dollar buying has paused, with no figure attached. Nothing here is a European-morning quote and the distances we have measured are distances from Tokyo closes, not from where London will start.
And we cannot tell you what the 12th is. A dollar–yen option expiring on 12 November, named this morning with no size, could be the largest structure on the board or a routine one. Six days is far enough out that it is irrelevant to today and near enough that it will be somebody’s reason for defending 157.80 next week.
Related
- FX event calendar — payrolls at 13:30 UTC, Canada in the same minute, and the New York cut 90 minutes behind both.
- Signals — what a published stop cluster is allowed to mean once price has already been through it.
- EA presets — position sizing calibrated on a 6.23 per cent volatility reading, going into a week priced at the high sevens.
