Three American events sit on today’s calendar and two of them are published by the Federal Reserve Bank of Chicago. The Chicago Fed prints its times in Central. One widely-read Monday round-up printed both of them with Eastern letters attached, and both landed in your schedule exactly sixty minutes early. The clock was never wrong. The two letters after it were.
What the issuer actually says
We went to the two pages that own these events rather than to a calendar, and both answer in one line.
- The Chicago Fed National Activity Index. The bank’s own data release calendar carries the August reading on 21 September and states the time as “8:30 a.m. ET/7:30 a.m. CT”. That is 12:30 UTC. The same page lists the September reading for 26 October at the same hour.
- Austan Goolsbee at OMFIF. The Chicago Fed’s own speech page for the engagement — an on-stage conversation in London hosted by the Official Monetary and Financial Institutions Forum — gives the start as “5:30 am CT”. That is 10:30 UTC.
Now the calendar we read at the London slot this morning. A Monday preview article on a large financial portal listed the activity index at “7:30 AM ET” and the Goolsbee remarks at “5:30 AM ET”. Convert those and you get 11:30 UTC and 09:30 UTC. Both an hour early. Both in the same direction.
The mechanism is the zone letters, and you can see it in the digits
Look at the numbers rather than the conversions. The Chicago Fed publishes the index at 7:30 Central. The round-up published it at 7:30 Eastern. The Chicago Fed publishes the speech at 5:30 Central. The round-up published it at 5:30 Eastern. In both cases the digits are copied across intact and the zone is changed.
This matters more than a generic “that vendor runs early” would. A random hour of error is unmodelable. A systematic one has a signature you can test for, and the signature here is specific: it only bites on events whose issuing institution is not on the East Coast. The Chicago Fed helpfully publishes both zones for its data release and only Central for its president’s diary, and both got flattened the same way.
We raised this pattern on Friday, when six American times on one calendar looked an hour early, and again this morning on these two events, where we declined to make it a rule on two instances. It is a rule now, and it is narrower and more useful than the one we were circling: when the issuer quotes a Central Time figure, check which zone your calendar stamped on it before you trust the window.
The bill auctions are the control, and they pass
The same round-up lists the three-month and six-month Treasury bill auctions at “10:30 AM ET”, with prior yields of 3.970% and 4.060%. That is 14:30 UTC and it is the standard slot the Treasury has used for bill auctions for years.
That is exactly what the theory predicts. The Treasury is in Washington, publishes in Eastern, and there is no Central figure sitting next to it waiting to be picked up by mistake. One page, three events, and the error appears on precisely the two with a Chicago clock behind them. If the fault were a global offset in the publishing pipeline, the auctions would be wrong too. They are not.
The same publisher’s newsroom contradicts its own preview
Here is the part that should stop you treating any single vendor page as the answer. The portal that printed “5:30 AM ET” in its Monday preview also covered the speech when it happened. That coverage is timestamped 06:35, updated 06:43, and it describes prepared remarks already delivered in London.
Read those timestamps in Eastern and they sit five minutes after a 10:30 UTC start — which is to say, five minutes after the Chicago Fed’s stated time and an hour after the preview’s. There is no reading of that timestamp consistent with a 09:30 UTC speech, because a newsroom does not file a summary of prepared remarks an hour before they are read. The conversion is ours and we are showing our work, but the direction is not in doubt: the newsroom and the issuer agree, and the preview is the outlier.
The portal’s own structured event page, incidentally, shows 12:30 for the activity index with no zone marker at all — the correct UTC time. So the publisher holds the right answer in one system and the wrong one in another, and which you get depends on whether you read the database or the prose. A third reader we checked, a large macro data site that renders in UTC, also gives 12:30.
One further note, offered as one fetch and nothing more: an exchange-hosted event page we opened for the same speaker returned an engagement from May rather than today’s. We are not building anything on that and we are not naming it as a pattern. We mention it because it cost us thirty seconds and it is the second metadata surprise of the morning.
What you do with sixty minutes
The directional question — does a regional Fed president talking about demand-driven inflation move dollar-yen — is not the useful one, and we are not going to answer it. The useful one is about exposure.
- If your news filter blocked 09:20 to 09:40 UTC, you were flat through nothing and live through the speech. That is the worst arrangement available: you paid the spread and the missed-trade cost of a filter and received none of the protection.
- The 12:30 UTC print is the one that is genuinely today’s event risk, and it is the same slot as every other 8:30 Eastern American release, so if your schedule already has a standing 12:30 window you are covered whether or not you knew the index was coming.
- The 14:00 UTC New York cut carries a 157.00 dollar-yen strike and a 1.1470 euro-dollar strike on today’s expiry board, published at 05:34 UTC. The notionals sit inside an image that does not extract, for the sixth time, so treat those as locations rather than magnets. Spot was quoted at 157.10–157.13 and 1.1484–1.1485 at 08:46 UTC, which puts both strikes close enough to matter and gives you no way to size the interest behind them.
- Audit the rest of your calendar for Central Time issuers. The Chicago Fed is the obvious one. So is the CME’s own scheduling, and so is anything sourced from a Chicago exchange. It is a one-off job and it is the only part of this note that is worth your Monday.
What this does not tell you
We have settled two times at the issuing institution and we have not settled the six from Friday. Those were American prints from Washington agencies, and the mechanism described here does not explain them — we do not have a second explanation and we are not going to invent one.
We have not established that the round-up gets Central Time wrong generally. We have two events on one page on one day, plus the control. That is enough to justify checking and not enough to justify an alert on every entry from that source.
The Eastern reading of the newsroom timestamp is our inference. The page does not print a zone next to it. We think the inference is forced, for the reason given, but it is an inference and if you need the speech time to the minute for a live decision you should take the Chicago Fed’s page and not ours.
And we did not watch the speech. What Goolsbee said — that demand may now be adding to inflation and that there is “no ambiguity” about the response if it overheats — reaches us through one report, and we have made no attempt here to price it.
