Six Federal Reserve and European officials have spoken about the rate path in the last seven days. Between them they have given a direction, a horizon, a risk list and, this morning, an argument against guidance itself. Not one of them has given a number. The only number in this desk’s file belongs to Governor Waller, who put the probability of at least one more increase by December at roughly 85 per cent on 8 October and then, speaking in New Delhi on 13 October, declined to repeat it. That is one cited probability across six speakers in six days — and in the same window the futures feeds this desk reads have carried October at four different values and December across a nineteen-point spread. The officials have stopped supplying the number. The vendors have not stopped supplying it, and they do not agree.
What the two presidents actually said, and what they carefully did not
The Boston president, speaking on 13 October from prepared remarks at a banking conference, said the Bank will likely need to raise rates once more before the end of this year, described it as a modest additional move, and tied it to returning inflation to the two per cent target in a timely and lasting way. She supported last month’s quarter-point increase, expects both headline and core inflation to run above two per cent in coming months, named jet fuel, diesel and home heating oil, and listed artificial-intelligence infrastructure spending, tariffs and Middle East energy costs as risks. She did not name a meeting and did not name a probability. Whether and when further action is needed, she said, depends heavily on how the incoming data shapes the outlook and the balance of risks.
The New York president, the same day, said one more increase is appropriate this year and that he currently sees no need for one in 2027. He said disinflationary forces are at play, that he sees no sign the labour market is adding to inflation, that he is not seeing broad price pressure outside artificial intelligence and energy, and that artificial intelligence is the inflation driver he watches most closely. He also did not name a meeting and did not name a probability. Three readers of each statement and not one of them reports a figure, which is as close to proof of absence as this material gets.
Set against Waller’s 8 October remark that the increases do not need to come at consecutive meetings, the shape is consistent and it is informative in its own way: three officials in six days have said the next move is coming and all three have refused to say when. That is not evasion. It is a reaction function being described honestly — a conditional, not a date.
Europe did the same thing before this slot published, and the ECB president said the quiet part out loud
This morning, inside the Tokyo session, two more officials added to the pile. The Belgian central bank governor said at 02:27 UTC that inflation remains too high and that another increase will likely be needed. The ECB president, between 02:33 and 02:45 UTC, said bond yields are rising worldwide and that she is watching medium-term inflation closely. Then she said the thing that explains the whole pattern: forward guidance is currently of little use.
Take that seriously rather than as a throwaway. If an institution judges that pre-committing to a path is unhelpful, the natural consequence is exactly what the last six days look like — officials stating a direction and a conditional and declining to attach either a date or a probability to it. A trader reading this as officials being coy has the causality backwards. They are not withholding a number they possess. They are telling you that a number of that kind is not the thing they are producing right now. Note the single-source caution: both of these remarks reach us through one Japanese vendor’s Tokyo-session news log, and this desk has one reader for each. We publish them as that vendor’s report rather than as settled wire copy.
The feeds are filling the gap, and the gap is not small
Through those same six days the futures-implied probability of an increase at the 27 to 28 October meeting has been read by this desk at 16, 19, 19.4 and 20 per cent across four separate feeds, and a Japanese broker this morning put it at above twenty per cent. December has been read at near 90, at over 80, at 80-plus and at 71 — a nineteen-point spread on a binary event two months out — and the same broker this morning gives roughly 90 per cent. The feeds are confident in a way that none of the officials has been.
The useful inference is not that the feeds are wrong; a futures market can price a distribution the speakers decline to summarise, and that is its job. The useful inference is about what you are actually trading when you trade a Fed headline. If the official communication contains a direction and no number, then the number in your terminal is a market construction, and a nineteen-point spread across readers of that construction means your error bar on it is wider than most of the moves it is used to explain. A system that treats an 85 and a 71 as the same input is not being robust. It is failing to notice that they are different numbers.
Today’s test is billed as supervision, so do not expect it to settle anything
This desk has a standing question: will any official repeat a probability figure? It has now scored negative five times running — Waller on 13 October, the two presidents on 13 October, and the two European officials this morning. The next scheduled test is the Vice Chair for Supervision at 09:40 UTC today, inside this slot’s window, and the Board’s own calendar files that appearance under the topic Supervision and Regulation, as it does her second appearance at 02:55 UTC tomorrow. On that billing a monetary-policy probability is unlikely, and a reader who positions for one has mistaken a supervision speech for a policy speech. The Richmond president at 17:30 UTC and the Beige Book at 18:00 UTC are the day’s other two Federal Reserve items, and both land after the print that actually matters.
Because the print is the event. US September consumer prices arrive at 12:30 UTC with headline expected at 3.6 per cent against a 3.4 prior and core at 2.5 against 2.4, per the broker reading cited below. Dollar–yen has spent the Tokyo session in a nineteen-pip range, 158.30 to 158.49 as at 01:00 UTC, twelve sen above a 158.33 New York close. Euro–yen is twenty-nine pips wide on the session and euro–dollar eight. Six central bank speakers have passed through the window from 02:10 to 08:00 UTC and the pair has moved less than twenty pips through all of them. That is the honest summary of what official talk is worth to this market today: the officials are not the event, the print is, and the market has already decided that.
What this does not tell you
It does not tell you the next increase comes in October rather than December, and nothing above should be read as a view either way; three officials declining to name a meeting is the opposite of evidence about which meeting. It does not tell you the feed probabilities are wrong — only that four readings of the October figure span four points and the December readings span nineteen, and that this desk has never reconciled them to a single method. It does not tell you the two European remarks are correctly reported: each has one reader, a single vendor’s session log, and a translation chain we have not independently checked. It does not establish that the Boston and New York statements drove any part of last night’s price action; this desk said yesterday that the overnight recovery from 157.83 has no cause we can source, and adding two speeches to the timeline does not change that. And it tells you nothing about the 12:30 UTC print, on which we take no view and hold no edge.
Related
- FX events calendar — today’s speakers and the 12:30 UTC print, in UTC.
- Signals — how a probability input with a nineteen-point spread should be weighted.
- EA presets — news-window and release-time configuration.
- Waller Spoke in New Delhi and Named No Probability — the 13 October piece that opened this tally.
- Friday’s Positioning Report Is Measured at Last Night’s Close — this morning’s companion on measurement instants.
