The Christian Democrats took 4.9% in Mecklenburg-Vorpommern on Sunday. That is the Chancellor’s own party, below the threshold, out of a state parliament. The Alternative für Deutschland took 38.2% and is the largest party in the state. Friedrich Merz called it a disaster and said he is staying. And EUR/USD, which traded 1.1477 to 1.1485 through the Tokyo hours, is doing what it did the last time this happened: nothing. The instrument that moved was the German front end.
The numbers, and the spread between the people counting them
Two German state elections resolved on Sunday 20 September, and you were not in the market for either of them. Here is what three readers have, and they do not quite agree, because some of these are exit polls and some are preliminary counts:
- Mecklenburg-Vorpommern. Danske Research, reading the preliminary count on Monday morning, has AfD 38.2%, SPD 35.5%, CDU 4.9%. Al Jazeera, reading Sunday night’s exit poll, has AfD 37% and SPD 36%. A third reader describes the AfD at roughly 38% and leading.
- Berlin. Danske has Die Linke 25.7% and CDU 18.8%. Al Jazeera’s exit poll has Die Linke 24.5%, CDU 20%, AfD 16%.
The gaps are a point or less and they are the ordinary gap between an exit poll and a count. We are not going to pretend that is a finding. The number that survives every reader is the one that matters: the AfD is the largest party in a German state parliament again, and in that same state the federal governing party did not clear five per cent.
Danske’s note adds that the state result is the CDU’s weakest in Mecklenburg-Vorpommern since the Federal Republic was formed, and that while a change of chancellor has become more likely, a breakdown of the federal coalition has not. That is one house’s political judgement and we pass it on as one house’s political judgement.
We have written this article before, and we were right for the wrong instrument
On 8 September this desk published a piece whose entire argument was that a record AfD result had moved the euro by nothing at all. Al Jazeera puts that earlier vote — Saxony-Anhalt — two weeks before Sunday’s. So this is the second instance of a thing we filed as an anecdote, and the honest scoring is split.
On the euro, the earlier piece holds. EUR/USD consolidated below 1.15 into Monday and was quoted at 1.1477 to 1.1485 through the Asian session. A chancellor’s party being ejected from a state parliament is worth, so far, less than a figure.
On the rest of the German curve, the earlier piece was not tested and would have failed if it had been. Danske has the two-year Bund up five to six basis points to 3.28% and the ten-year up five to 3.52% on the result, and describes a political-risk-premium narrative attaching itself to Germany. That is a single reader and we label it as one; nobody else we read this morning priced the move. But five basis points at the front of the German curve on a Sunday state election is not nothing, and a desk that published “this does not move” three weeks ago should say plainly that it was looking at the wrong screen.
The same note has EUR/SEK at one-year highs around 11.30 and EUR/NOK unchanged around 10.80. If you want the euro-area political risk premium, it is being expressed against the Scandinavian crosses and the Bund, not against the dollar.
What you actually do with a Sunday event
The trade here is not directional and we are not going to pretend to one. It is a scheduling question, and it has an uncomfortable answer.
A German state election is not a field on your economic calendar. It has no consensus, no prior, no release time in the sense your news filter understands, and it resolves on a Sunday when you cannot hedge it and cannot trade it. By the time your system sees a price on Monday, the repricing has already happened in an instrument most FX systems do not watch. If you were short EUR/SEK into the weekend you have been marked against a political event you were not tracking and could not have exited.
So the question is not “which way does the euro go”. It is whether your weekend gap risk is sized for political events you do not have on the calendar at all — and whether the correlated exposure you carry into a Friday close is measured against the Bund as well as against EUR/USD. For most retail systems the answer to both is no, and the cheapest fix is a position-size rule on Fridays rather than a new data feed.
The rest of the European morning is quiet, and that is the risk
There is no euro-area data today. The European Central Bank’s president speaks on Tuesday at 11:00 UTC, and the flash purchasing managers’ indices for the euro area land Wednesday at 08:00 UTC with the United Kingdom’s at 08:30. Christine Lagarde’s most recent framing, as reported this morning, is that the Bank decides meeting by meeting and could hold, raise or cut, with cuts “very unlikely at the moment”. That is not a signal, it is a refusal to give one, and it is worth reading as such.
Which means the London open at 07:00 UTC is the first real price discovery on Sunday’s result, into a market where Tokyo is shut for a third day and will be shut tomorrow and Wednesday too. Thin books and a political headline with no scheduled resolution is the configuration that produces moves nobody can attribute afterwards. Expect the range, not the trend, and do not size as though the first print after the open is information.
What this does not tell you
- The Bund move is one reader. The 3.28% and 3.52% levels and the five-to-six basis point moves come from a single house’s morning note. We did not find a second reader who priced the election into German yields, and we did not fetch the yields independently. If that number is load-bearing for you, check it yourself before you act on it.
- We do not know these are final results. Every figure above is described by its own source as an exit poll or a preliminary count. State results are certified later and can move by tenths.
- We cannot tell you the euro is not moving because of the election. A quiet EUR/USD on a Monday with no euro-area data and Tokyo shut is consistent with the election being ignored and equally consistent with there being nobody there to trade it. Those are different worlds and today’s tape does not separate them. Wednesday’s flash PMIs, with Tokyo back and a real calendar, will tell you more than today will.
- We have no coalition arithmetic. None of the three readers gave us seat counts or turnout, so we cannot tell you who governs either state, and the governing question is the one that eventually reaches the Bund.
- Danske’s chancellor-change judgement is a political forecast, not a market measurement. We report it because it is attributable and specific. It is not evidence.
Related
- FX event calendar — scheduled releases, and a reminder of what it does not contain
- Signals
- EA presets — where weekend and news-window sizing rules live
