Moody’s left France at Aa3 on Friday evening, after the New York close, with the outlook still negative. The bond market had priced the other outcome: two readers say most participants expected a one-notch cut, which would have aligned Moody’s with Standard and Poor’s and Fitch, both of which downgraded France after the 2024 snap election. The interesting part is not the hold. It is the sentence Moody’s hung it on — that its Aa3 assumes a deficit-reducing 2027 budget will be adopted — because the lower house spent last week removing savings measures from exactly that budget. The rating was held on a condition the legislature is actively testing, and Moody’s added that the review was not a formal rating action at all.
What was actually decided, and what was not
Aa3, unchanged. Outlook negative, where it has been for a year. Announced after the New York close on 23 October, which is why a Tokyo desk is the first to price it.
Moody’s stated reasoning, as two independent readers report it: slower growth and wider deficits have pushed the debt burden somewhat higher than previously expected; the sharp rise in bond yields heightens the risk of a further weakening in debt affordability; a downgrade would likely follow additional evidence that legislative institutions’ capacity to contain that deterioration has weakened further. One reader adds that Moody’s expects financing costs and social tensions to ease somewhat in coming months, and that the Aa3 assumes a deficit-reducing 2027 budget is adopted.
And one reader reports Moody’s saying the review was not a formal credit rating action, with no indication whether one would follow soon. That is a procedural fact and it matters more than it reads. If this was a periodic review rather than an action, then the thing the market spent the week pricing was not scheduled to produce a decision in the first place, and the absence of a downgrade is weaker news than “France avoids downgrade” makes it sound. We are reporting that distinction because one publisher drew it and the others did not.
The condition and the counter-evidence arrived in the same week
The 2027 budget targets a deficit of 5 per cent of output, down from 5.4 per cent this year. One reader reports that the lower house has scrapped some key savings measures, that opposition parties are resisting compromise, that the government may struggle to pass the bill, and that it could face a no-confidence vote. The same reader reports violent protests over school funding escalating into riots during cost-of-living demonstrations, with further protests called for Saturday 24 October, and a government refusing both more school funding and a minimum-wage rise.
So Moody’s names two things it expects to improve — financing costs and social tensions — and conditions the rating on a budget passing. Three assumptions, and the week’s news ran against all three. We are not saying Moody’s is wrong; a rating agency forecasting a year out is entitled to look past a fortnight. We are saying that if you are trading the downgrade risk rather than the rating, the hold did not reduce it. The negative outlook and the stated trigger — additional evidence on parliament — are both still in place, and the French legislative calendar is where the evidence comes from.
One analyst quoted by the same reader goes much further, arguing that a joint cut to BBB by Moody’s, Standard and Poor’s and Fitch could force selling by investors with rating limits. We note it and we do not endorse it: BBB is many notches below Aa3 and that is a scenario, not a forecast. It is in here because it is the mechanism by which a rating becomes a flow rather than an opinion, and because it is the version of this story a system trader should have a sizing rule for.
Our own published band was too narrow, by one basis point
On 22 October this desk published that four readers put the French ten-year between 5.076 and 5.13 per cent after it went above 5 per cent for the first time since 2002. A fifth reader, reached this run, puts Thursday’s high at 5.14 per cent and calls it the highest since July 2002. That is 1.0 basis point above the top of the band we published and 6.4 above the bottom.
One basis point is immaterial to a position and material to the claim we made. We did not publish a level; we published a band and implied it was the range of what anybody says. It was not. The honest version is that five readers now span 5.076 to 5.14, a 6.4 basis point spread on the single number the whole French story is quoted through, and that the band widens every time we read another publisher. The same reader also gives the premium over Bunds as the highest in fifteen years on 22 October, which is consistent with the widest-since-2011 framing we published and with our arithmetic that 174.6 basis points is 77.6 per cent of the November 2011 peak of 225.1.
There is a second item here we read on Friday and did not publish, and it belongs to this desk because this desk published the decomposition that depends on it. One publisher’s single analysis page gave the German ten-year as 3.34 in one section, 3.32 in another and 3.31 in a third, alongside a note that it was down from 3.41. Our 22 October article attributed 15.3 per cent of the spread move to the German leg. A 3 basis point internal spread inside one document, on that leg, is not enough to overturn an 84.9 / 15.3 split — but it is enough that we should stop quoting the split to one decimal place, and we are saying so rather than leaving it in a file.
Where a French rating reaches a yen position
One Japanese research desk, writing this morning, describes Japanese institutional selling of French government bonds as a euro-selling and yen-buying factor. That is a single source and a characterisation rather than a measurement — nobody publishes the flow — but it is the channel, and it is the same channel this desk has carried unread for several weeks and then watched arrive as a yen bid on 22 October.
The reason to keep it in view this week is the calendar rather than the rating. Dollar–yen closed Friday between 156.86 and about 157.08 depending on which publisher you read, after a New York session that fell to 156.31 and recovered roughly 77 pips. The week holds the FOMC decision at 18:00 UTC on 28 October with Chair Warsh at 18:30, the ECB at 13:15 UTC on 29 October with Lagarde at 13:45, and the Bank of Japan around 03:00 UTC on 30 October with Governor Ueda at 06:30. All three are expected to hold. If the French leg moves the euro against the yen inside those windows, it arrives as a cross flow on a day you are positioned for a central bank, which is a sizing problem and not a view.
One correction to anybody’s stored calendar, and it is the reason we re-derived those hours rather than carrying them: Europe left summer time yesterday. The ECB’s own weekly calendar still says 14:15 CET for the decision and 14:45 for the press conference, exactly as it said before — but CET is UTC plus one, so the UTC hour moved from 12:15 to 13:15. The local time did not change and the UTC time did. American hours do not move until 1 November.
What this does not tell you
We did not reach Moody’s own release. Every figure and every quoted intention above is a secondary reading of it, from two publishers that agree on the rating, the outlook, the date and the broad reasoning and differ on which parts of Moody’s statement they report. The not-a-rating-action detail is single-sourced. So is the 2027 deficit target, so is the no-confidence risk, and so is the Japanese-selling channel.
Nothing here is a view on the euro, on French paper or on the yen. We cannot tell you whether the hold is priced, because we read no post-announcement European pricing — the decision landed after the New York close and the first market to express an opinion on it is the one opening now. We also cannot tell you what the OAT did on Friday; the 5.14 is Thursday’s high, and we refused every untimestamped level on the page that carried it. If you want one number from this, it is that our own published band on the single most-quoted yield in Europe was wrong at the top by a basis point, and we found that by reading a fifth publisher rather than by anything happening in the market.
Related
- FX event calendar — the FOMC, ECB and BOJ windows this week, in UTC
- Signals
- EA presets
