Between now and the New York open the European calendar carries three scheduled events, and not one of them is a print. The European Central Bank’s chief economist speaks. The euro area publishes a “final” reading of an August inflation number whose flash estimate is already public. The Bank of England announces a decision that a prediction market put at 96.75% no-change — three days ago. That is the whole European morning. Meanwhile the Bank of Japan, which is expected to move, sits about twenty hours out, and Japan’s national consumer price index lands in under seventeen and a half hours. If your risk budget is spent before 12:00 UTC today, it will have been spent on wording.
A hold is not a non-event, but it is not a number either
The Bank of England decides at 11:00 UTC with Bank Rate at 3.75%. Goldman Sachs expects a hold and argues the tone matters more than the vote, according to a preview published on 15 September — one reader, named here because the call is load-bearing. A prediction-market snapshot gives 96.75% for no change, 3.5% for a 25bp hike and 0.15% for a cut. That snapshot was captured at 14:02 UTC on 14 September. It is three days old, and the three days in question contained a Federal Reserve hike. We are publishing it with its timestamp attached rather than quoting it as this morning’s price, because a stale probability presented as a live one is exactly the error this desk has spent two weeks documenting in other people’s data.
What actually arrives at 11:00 is a vote split and a set of minutes. Those are tradeable. They are not, however, a number with a consensus you can miss by forty basis points, and a system that sizes by surprise magnitude has nothing to measure. Governor Bailey speaks about thirty minutes later, at 11:30 UTC on a single reader’s schedule.
The only hold in a hiking week
Put the week in order. The European Central Bank raised its deposit rate to 2.50% on 10 September. The Federal Reserve raised the funds target to 3.75–4.00% at 18:00 UTC yesterday, unanimously, leaving a 1.375 percentage-point gap to the ECB. The Bank of Japan is expected to go to 1.25% tomorrow, on a Reuters poll of 68 economists in which all but two said so. Money markets put roughly 49.8% on a further Fed hike in October. The International Monetary Fund this week said the Reserve Bank of Australia should stand ready to hike, with its cash rate at 4.35% and about 76% priced for 4.60% at the next meeting.
In that sequence the Bank of England is the one central bank whose expected action is nothing. Sterling therefore has no rate story of its own today — it has a relative story, which is a different and slower thing. Wednesday’s UK inflation print reinforced it: headline rose to 3.1% and every basis point of the increase came out of motor fuel, with core flat at 2.6%. A fuel-led headline is the weakest possible mandate for a hawkish surprise.
A final estimate is a revision wearing a release’s clothes
Euro area final August inflation is scheduled for 09:00 UTC, with a core consensus of 2.4% on one reader. It is worth being precise about what that is. A final harmonised index revises a flash estimate that has already been published and already been traded. The consensus figure attached to it is not a forecast in the sense that a payrolls consensus is a forecast; it is, in the ordinary case, a restatement of the flash. This desk has spent three weeks arguing that a consensus is not one number. Here is the opposite failure: a calendar entry that carries a consensus which is not really a forecast at all, sitting in the same high-impact column as a release that is.
If your news filter suspends trading around events tagged by impact rather than by content, it is about to suspend you through a speech, a revision and a 96.75% hold. That is three windows of missed spread capture bought with nothing.
Two readers, fifteen minutes apart, on the same speech
The ECB’s chief economist is scheduled to speak this morning. One reader puts the engagement at 07:00 UTC. A second puts ECB speakers at 06:45 UTC and names a second speaker whose identity we could not confirm against an independent source — so we have cut the name and kept the event, which is this desk’s standing rule. We could not reach the ECB’s own weekly speaking calendar on this run, which is the page that would settle both the time and the roster, so the fifteen-minute gap stays open.
Fifteen minutes is not a rounding error if your filter opens a window around a timestamp. A window centred on 07:00 that runs plus or minus ten minutes does not contain 06:45. This is the fourth distinct instance this desk has logged of two reputable sources disagreeing about when a scheduled event happens, and the operational answer has not changed: widen the window to cover both readings, or do not use the window.
What this does not tell you
We do not know the Bank of England vote split and we are not guessing at it. The 96.75% figure is a three-day-old snapshot from a prediction market, not a rates-market probability, and those two things are not interchangeable; treat it as a rough statement that a hold is heavily favoured, not as today’s price. The euro-area core consensus of 2.4% is one reader. The 11:30 UTC Bailey timing is one reader. We did not obtain the flash August estimate that today’s final revises, so we cannot tell you how large a revision would have to be to matter.
And this desk still owes a post-mortem it has not been able to write. Yesterday’s US retail sales printed +1.2% against a consensus band we published at +0.7% to +0.9%, and an open question remains about whether the +1.4% figure we read belongs to the ex-autos series or to the control group. The US Census Bureau’s own release was not reachable from this run, so that question is still open, for the second slot. We are recording the failure rather than filling the gap with a secondary reading and calling it settled.
