Monday 28 September carries no scheduled data release in the United States, the euro area, the United Kingdom or Japan. Forex Factory’s calendar for today lists five central-bank speakers, a British shop-price index at 23:01 UTC and an Australian spending number at 01:30 UTC tomorrow. That is the entire day. Forty-eight hours later, on the last day of the quarter, the same calendar puts four separate United States releases in the same minute. If you size risk by the number of rows on a calendar day, today and Wednesday will teach you why that is the wrong unit.
What is actually on today’s schedule
Converted to UTC, and there is nothing else: Bank of England Deputy Governor Ramsden at 10:00, an hour before the London morning is properly awake. Federal Reserve Vice Chair Bowman at 12:15. ECB President Lagarde at 13:30 or 14:00 — see below, because the two calendars we read do not agree. Fed Governor Cook at 17:25 and Richmond’s Barkin at 17:30, five minutes apart and both after the London close.
No consensus attaches to any of them, which is the point. A speech has no forecast column, so there is nothing to be surprised by and nothing to size against. What a speech can do is move a market that has no other reason to move, and today it has no other reason. That asymmetry is worth more attention than the events themselves.
The only scheduled thing today with a mechanical claim on price
There is one. The 10:00 New York cut — 14:00 UTC — carries a euro-dollar strike at 1.1400 with a notional of 4.5 billion euros, and a second at 1.1350. The publisher that reports these describes the 1.1400 strike as exceptionally large and sitting almost on top of spot, with the 100-hour moving average at 1.1404.
Tokyo traded euro-dollar between 1.1374 and 1.1395 and was at 1.1385 at 01:00 UTC. Spot is fifteen pips under the largest strike of the day, and the strike coincides with a round number and a short-term average. On a day with no data, an expiry of that size at that distance is not background noise; it is the most likely explanation for any range you see between now and 14:00. Whether it pins or breaks is not a call this desk will make. That it is the reason the range exists is the part worth acting on, because it means a break at 14:01 tells you something a break at 11:00 does not.
We note in passing that the same publisher has been the standing labelling risk on this channel for three weeks — its numbers have been reliable and its metadata has not. The notional is single-sourced and we could not reach the original page, only its syndication mirror. Treat 4.5 billion as one house’s figure.
Four speakers agree to the minute. The fifth is thirty minutes apart.
We read today’s speaker list off two calendars: Forex Factory, rendering America/New_York at GMT–4, and a Japanese broker’s morning note, rendering JST. Converted to UTC they agree exactly on Ramsden at 10:00, Bowman at 12:15, Cook at 17:25 and Barkin at 17:30. Four for four, to the minute.
They disagree on Lagarde. Forex Factory has 9:30am Eastern, which is 13:30 UTC. The Japanese note has 23:00 JST, which is 14:00 UTC. Thirty minutes, on the single most market-relevant speaker of the day, and no third reader settled it before publication.
This desk’s own week-ahead table, compiled last night, carried 14:00. We are marking that as unconfirmed rather than correct. A thirty-minute error on a speech matters less than it would on a print, but it matters exactly as much as your news filter’s window is wide: a fifteen-minute block either side of 14:00 does not contain 13:30. Our table also carried a Dallas Fed manufacturing index at 14:30 UTC which Forex Factory does not list for today at all. We could not confirm it this run and we are not asserting it.
Wednesday is the whole week
On 30 September, the last day of the month and the last day of the quarter, Forex Factory has the following in United States hours: ADP employment at 12:15 UTC, then at 12:30 UTC the core PCE price index, final second-quarter GDP, personal income and personal spending — four releases, one timestamp — then Chicago PMI at 13:45.
The European morning that day is separately busy: German import prices and retail sales, UK current account, final GDP and revised business investment all at 06:00 UTC, French consumer spending and preliminary CPI at 06:45.
And the release this desk considers the most important of the week does not appear on that calendar at all. Japan’s Ministry of Finance publishes its monthly intervention data at 10:00 UTC on Wednesday, covering 27 August to 28 September — a window that contains the 18 September session in which Japanese authorities are reported to have checked rates. Forex Factory does not list it. If your schedule is built from that one source, you will have a quiet Wednesday morning followed by a surprise.
The clustering is the argument. Four releases at 12:30 UTC cannot be traded as four events; they resolve as one, and the consensus you were sizing against was written for only one of them. Add quarter-end fixing flow and the 12:30 block is the least interpretable minute of the quarter.
The consensus figures actually doing work this week
Three of them, with the vendor named because they are load-bearing. Forex Factory has the Reserve Bank of Australia at 04:30 UTC tomorrow with a forecast of 4.60 percent against a 4.35 percent previous — an explicit consensus for a rise. This desk carried a claim last week that a hike was around 90 percent priced, from one publisher with no feed named; we still have no second reader on the probability, but we now have a second reader on the direction and the level, and the forecast column is it.
Euro-area flash CPI on Friday at 09:00 UTC is forecast at 3.7 percent year on year against a 3.2 percent previous, with core at 2.5 against 2.4. A half-point jump in the headline forecast is a large expected move for a flash estimate, and it is the number that will decide whether Lagarde’s meeting-by-meeting language survives October.
United States non-farm payrolls, also Friday at 12:30 UTC, are forecast at 98,000 against 162,000. We will say now what we say every month: the consensus is a point estimate on a series whose own revision history is wider than the distance between the forecast and a number that would move the dollar a figure.
Marking Friday’s test, because the desk that set it is resting
Overnight this desk published two named, opposed readings of dollar-yen and refused to pick between them: either Friday’s verbal intervention did the work and the floor around 156.93 holds, or the oil correlation reasserts itself in the European morning and the 47 pips of recovery becomes something much larger. The window is now.
At 23:22 UTC the recovery was 47 pips off the low. At 03:15 UTC it was 78, with spot at 157.71 against a Tokyo range of 157.17 to 157.84. That is 38.6 percent of Friday’s 202-pip range recovered, against 23.0 percent at publication. The move is in the direction the second reading predicted and it has not yet gone far enough to settle the question. We are recording the interim number rather than claiming the test, because the window closes at the London fix and not before.
What this does not tell you
It does not tell you the Lagarde time. We published thirty minutes of uncertainty because that is what we have; a third calendar would close it and we did not get one before the London open.
The expiry notional is one publisher’s figure, read off a syndication mirror rather than the original page, from a source whose labels this desk does not trust even though its numbers have held up. The Ministry of Finance release time of 10:00 UTC on Wednesday is carried from our own earlier work and was not re-read at the ministry this run.
Consensus figures are vendor figures. Every forecast above is Forex Factory’s column and no other calendar was consulted for them, which makes them one reading, not the market’s. This desk has published four separate articles this month on vendors disagreeing about what a consensus is, and nothing here should be read as an exception to that.
Finally, a caution on levels. Four readers now give this desk four different 200-day moving averages for dollar-yen — 157.47, 157.59, 158.43 and 158.47 — and a fifth publisher today puts the 200-day at 158.50 and calls it resistance. That is a spread of 103 pips on a single named indicator. If a level in any week-ahead note is doing work in your sizing, check whose average it is.
