For eight slots this desk has written the same sentence about option expiries: the strikes are in the text, the notionals are inside a picture, and a strike without a size is a location rather than a force. This morning the picture was unnecessary. Two separate publishers put the full 16 October ladder out in plain text, with notionals, before the London session was an hour old. The field we could not read is readable. It is also not one number. On four of the eight rows the two houses share, they disagree — by 200 million euros at one strike, by 400 million at the next, by 100 million dollars on the dollar–yen row, and on the Swiss franc row by 35 pips of strike while agreeing on the size to within two million. You now have two hours and eight minutes to the New York cut and a notional ladder whose largest single entry is 21 per cent bigger at one house than at the other.
Four rows identical to the million, four rows not
Take the two published tables side by side. One is stamped 06:21 UTC, the other 09:55, with a companion post from the same house at 10:04 carrying figures identical to its sibling — so this is not one house typing twice and getting two answers. It is two houses reading the same market and publishing different sizes.
Where they agree, they agree exactly: euro–dollar 1.1160 at 765 million euros, dollar–yen 157.00 at 680 million dollars, Australian dollar 0.6950 at 714 million, euro–sterling 0.8450 at 749 million. Four rows, four exact matches, down to the last million. That is not two estimates converging. That is two readings of a common source.
Then the other four. Euro–dollar 1.1200: 1.30 billion euros at one house, 1.1 billion at the other — 200 million apart, 18.2 per cent of the smaller figure. Euro–dollar 1.1225: 2.30 billion against 1.9 billion — 400 million apart, 21.1 per cent. Dollar–yen 158.00: 925 million dollars against 825 million — 100 million, 12.1 per cent. And one house carries a fourth euro–dollar strike at 1.1350 for 531 million euros that the other does not list at all.
Add the euro–dollar column up and the two ladders differ by 1,131 million euros on totals of 4,896 and 3,765 — exactly 30 per cent of the smaller. That is the number to sit with. The same instrument, the same expiry date, the same cut, two publishers, and a third of the book in dispute.
The two strikes the morning’s pin argument rested on are 600 million apart
At 01:23 UTC a third publisher put out the page this channel has used all week. It named two euro–dollar strikes, 1.1200 and 1.1225, said they sandwiched spot within fifteen pips either side, and flagged a pinning effect into the cut. Its own notional table was the image. We wrote about that page at 06:18 this morning and attacked its unstated spot premise. We did not attack the ladder, because we could not see it.
We can see it now, and there is a second problem. Those two strikes carry 3,600 million euros between them at one house and 3,000 million at the other — 600 million apart, 20 per cent. The ratio between the two strikes survives better than the levels do: 1.769 at one house, 1.727 at the other. So both tables agree that the upper strike is roughly three-quarters bigger than the lower one, and disagree about how big either of them is.
If you size an expiry trade off a notional, that matters more than it looks. A gravitational argument is a function of size against liquidity. Twenty per cent of the stated size is not a rounding difference; it is the difference between an expiry that is worth building a session around and one that is background.
And the page that produced the pin showed two rows of a four-row ladder
The 01:23 page said the two strikes were “both” for euro–dollar. On the fuller tables there are four euro–dollar strikes at one house and three at the other. The 1.1160 at 765 million sits below the band. The 1.1350 at 531 million sits above it, on the table that carries it.
That is an editorial selection, not an error, and the two largest strikes are a defensible thing to lead with. But a reader who took that page as the ladder had the band and neither shoulder, and the shoulders are what tell you whether price leaving the band has anywhere to be pulled to. Today it does: 84 pips above the current level, at a strike one of the two houses does not print.
Meanwhile the band is not containing anything. Euro–dollar was 1.1226 at 06:30 GMT, a pip above the upper strike. At 08:21 one reader had it rising toward 1.1250. At 10:54 another had 1.1266 — 41 pips above the upper strike, 66 above the lower, and 40 pips of travel in four hours and twenty-four minutes. Price did not get pinned into the European morning; it walked out of the top of the band and kept going while the dollar index made a two-week low.
The Swiss franc row is the one that proves the point
One table gives dollar–Swiss at 0.8285 for 560 million dollars. The other gives 0.8250 for 558 million. The strikes are 35 pips apart. The sizes are two million apart — 0.36 per cent.
Two options with nearly identical notionals and materially different strikes are not two different options that happen to look alike. Almost certainly this is one option, with one of the two publishers carrying a wrong digit in the strike field. Which one is wrong is not determinable from here, and we are not going to guess.
That is the whole finding in miniature. Four rows match perfectly, which says there is a common upstream feed. Four rows differ, which says at least one publisher is transforming, filtering or mistyping on the way out. And on the row where the error is most obviously an error, the disagreement is in a field both houses would tell you is simply copied.
What to do with a ladder you cannot second-source
The instruction is not “stop reading expiry pages”. It is narrower than that.
Store the publisher and the hour with every notional, the way you should already be storing them with every level. A notional is not a market observable you looked up; it is a number a house published at a time, and this morning two houses published nine rows and agreed on five.
Where the two readings bracket, size to the smaller. Three billion euros across the two strikes rather than 3.6 billion, 825 million dollars at 158.00 rather than 925. If the trade only works on the larger figure, the trade is sized off a disagreement.
And treat the single-source rows differently from the agreed ones. The 1.1350 strike exists on one table and not the other. That is not a bigger or smaller version of a fact; it is a fact one reader does not have. Today it is the nearest strike above spot, which is exactly the row a price walking up out of the band would care about, and it is the row with the weakest provenance in the whole ladder.
You have two hours and eight minutes until 14:00 UTC. Import prices land at 12:30 and industrial production at 13:15, both inside that window, both confirmed at the issuing agency rather than at a calendar. If you were planning to express a view on where the cut prints, the honest version of that view now has an error bar on its own inputs that nobody published this morning and that we only have because two houses happened to disagree in public.
What this does not tell you
We have not read the underlying option book and cannot. Everything above is two publishers’ descriptions of a market neither of them makes, and the agreement on four rows is evidence of a shared source rather than of accuracy — a common feed can be wrong in both copies at once.
We do not know which house is right on any disputed row, and we have deliberately not split the difference. An average of two numbers that disagree is a third number nobody published.
The 1.1350 strike appears on one table only. We have published it because it is the nearest strike above the current level and a reader deserves to know it exists, and we have labelled it single-sourced here rather than letting it sit in a list as though it had the same standing as the 765 million at 1.1160.
The spot levels in this piece come from three different readers and the spread between them is live, not historical: at roughly the same hour one timed reader gives euro–dollar at 1.1266 and an untimed ticker gives 1.12565, 9.5 pips apart, while on dollar–yen the same two sources are 18 pips apart against a one-pip gap between two timed readings four hours apart. We have used the timed figures and said where the untimed one sits.
And none of this is a view on direction. Price is above the band with the dollar index at a two-week low and two American releases still to come. Whether it stays there is not something an expiry ladder answers, and it is not something this desk is telling you.
Related
- Event calendar — the 12:30 and 13:15 releases and the 14:00 cut, in UTC.
- Signals — how we treat a published level whose source disagrees with itself.
- EA presets — news-window and sizing settings for a release block inside an expiry window.
- Two Level Sets, One Cross, Sixty-Seven Minutes, Nothing in Common — this morning’s piece, which attacked the same expiry page’s spot premise without being able to see its ladder.
- Three Positioning Numbers Before the 12:30 Block, All Broken Differently — where we first wrote that the notionals were inside an image.
