In short
- Best performing currency: the Canadian dollar, +1.19% against the basket of the eight majors, from the London opening on 1 October to the one on 2 November 2026.
- Worst: the euro, -1.56%. The biggest pair move of the 28 was EUR/CAD at -2.72%.
- Central banks: five of the eight met. The Bank of Canada raised to 2.50%; the Fed, the RBNZ, the ECB and the Bank of Japan held.
- Our setups with the mark: 29 calls, 15 the expected way, -3.32% together. The ones without it, in the same lists: 13 of 27, -0.92%.
- The ranking: the currency we ranked strongest beat the one we ranked weakest on 11 of 19 mornings, but lost -1.46% against it added up.
Best and worst performing currencies of the month
| Currency | October against the basket |
|---|---|
| Canadian dollar | +1.19% |
| Swiss franc | +0.88% |
| British pound | +0.70% |
| Australian dollar | +0.41% |
| New Zealand dollar | -0.35% |
| Japanese yen | -0.39% |
| US dollar | -0.85% |
| Euro | -1.56% |
The month was the euro going down against almost everything. All six of the biggest pair moves had the euro or the US dollar on the losing side: EUR/CAD -2.72%, EUR/CHF -2.42%, EUR/GBP -2.25%, USD/CAD -2.02%, EUR/AUD -1.97% and USD/CHF -1.71%. The quietest pair was NZD/JPY at +0.04%. The Swiss franc finishing second is the odd one out: its policy rate of 0% is the lowest of the eight, and it still ended the month as one of the strongest.
The first half: risk came back, then the data turned
October opened in risk off. Our risk sentiment reading, the heaviest of the four pillars at 40, stood at -20 on 1 October. On 2 October two releases pulled the dollar and the euro in different directions. Euro area inflation came in at 3.8% for September against 3.6% expected, with energy up 18.8% (Eurostat). The US jobs report showed 29,000 new payrolls against about 90,000 expected and unemployment at 4.2% (BLS).
Then stocks rallied. The reading crossed into risk on by 7 October, at +27, and peaked at +31 on the 8th. The S&P 500 closed at a record on Friday the 9th. The Australian dollar led the basket that week. The same Friday Canada reported 68,000 jobs lost in September, the second monthly fall in a row (Statistics Canada), and the Canadian dollar was the weakest of the eight that day.
The turn came in the week of 12 October. A Houthi missile hit Riyadh airport on the 10th, the third attack that week, and oil kept climbing: WTI went from around 88 dollars on 7 October to around 95 dollars a week later. US inflation on the 14th came in at 3.7% for September against 3.6% expected, core at 2.6% (BLS), and stocks fell that evening. Overnight Australia reported 39,300 jobs lost and unemployment up to 4.7% (ABS). On 16 October the VIX went above 21 and the S&P 500 fell 1.0% on the day.
The second half: oil, the Bank of Canada, and a week of five banks
Our reading was at -21 on 20 October and stayed below zero or close to it for the rest of the month, ending at +15 on the 30th. Oil did not come down. WTI was around 102 dollars on 28 October and around 104 dollars on the 30th. The Canadian dollar, the currency of an oil exporter, and the franc, one of the havens, both did well in that mix.
The last week carried five rate decisions in three days. The RBNZ held at 2.75% on 28 October against expectations of a hike, on a 4 to 3 vote according to market reports, and the New Zealand dollar fell sharply. The Bank of Canada raised by a quarter point to 2.50% the same day and said further hikes may be needed, according to its press release. The Fed held at 3.75 to 4.00% that evening, with Lisa Cook and Beth Hammack voting for a hike (Federal Reserve).
On 29 October the ECB held at 2.50%, unanimously, after discussing a hike (ECB), and announced that Christine Lagarde steps down on 7 February 2027 (ECB). US growth for the third quarter came in at 3.4% annualised and PCE inflation at 3.6% for September (BEA). On the 30th the Bank of Japan held at 1.25% by 8 to 1, with Takata voting for 1.5% (Bank of Japan), German inflation eased to 2.8% for October (Destatis) and Nasdaq stocks fell 2.5% on the day.
Strong American data and a Fed with two votes for a hike did not lift the dollar. It ended the month seventh at -0.85%. Every release of the month, with its time and source, is on the calendar.
All eight central banks, one table
| Bank | Rate | Last decision | Next meeting |
|---|---|---|---|
| Bank of Japan | 1.25% | Held, 30 Oct, 8 to 1 | 18 Dec |
| Federal Reserve | 3.75 to 4.00% | Held, 28 Oct, 10 to 2 | 9 Dec |
| European Central Bank | 2.50% | Held, 29 Oct, unanimous | 17 Dec |
| Reserve Bank of New Zealand | 2.75% | Held, 28 Oct | 9 Dec |
| Bank of England | 3.75% | Held, 17 Sep, 6 to 3 | 5 Nov |
| Reserve Bank of Australia | 4.60% | Raised 0.25, 29 Sep, unanimous | 3 Nov |
| Swiss National Bank | 0% | Held, 24 Sep | 10 Dec |
| Bank of Canada | 2.50% | Raised 0.25, 28 Oct | 9 Dec |
The one bank that raised was the bank of the best performing currency, and the one hold that surprised the market was the bank of a currency that finished in the bottom half. The Fed and the ECB both held as expected, and both their currencies finished in the bottom two. Every rate, with the date of the next meeting, is on the central banks page.
What the research said, and how it went
19 mornings were measured in October. The setups that carried our mark are in the table at the top of this page: 15 of 29 the expected way, -3.32% together. The ones in the same lists without the mark did 13 of 27 for -0.92%, and all of them together came to 28 of 56 and -4.24%. In September the mark had made +3.75% while the rest lost. In October it was the other way round, and that is the honest headline of the month.
The euro was read right. Short EUR/USD carried our mark on two mornings and gained on both, +0.57% together, the best setup of the month. Long USD/JPY added 2 of 3 for +0.27%.
The franc and the Australian dollar were read wrong. Long USD/CHF was marked on nine mornings, more than any other pair, and went 4 of 9 for -1.56%, while the franc climbed to second in the basket. Long AUD/CHF went 4 of 8 for -0.95%. The worst single morning was 14 October, when Australian jobs fell overnight: long AUD/CHF lost -1.39% and long AUD/JPY -1.34% from that opening to the next.
A month is a small sample, and 29 calls is still not many. But this is the first month with the mark on the site from the first day to the last, and it lost. Every morning, with the setups that did not carry the mark one switch away, is on the track record.
Did the ranking itself work?
The currency we ranked first beat the one we ranked last on 11 of 19 mornings. That is a small majority, and it hides that the misses were bigger than the hits: added up, the first did -1.46% worse than the last. Holding the ranking of 1 October to the opening of 2 November shows why.
| Ranked on 1 October | Score | 1 October to 2 November | Finished |
|---|---|---|---|
| 1. US dollar | 69 | -0.85% | 7th |
| 2. Japanese yen | 55 | -0.39% | 6th |
| 3. Australian dollar | 54 | +0.41% | 4th |
| 4. British pound | 53 | +0.70% | 3rd |
| 5. Swiss franc | 50 | +0.88% | 2nd |
| 6. Euro | 50 | -1.56% | 8th |
| 7. New Zealand dollar | 49 | -0.35% | 5th |
| 8. Canadian dollar | 43 | +1.19% | 1st |
The two ends were upside down. The dollar we put first finished seventh, and the Canadian dollar we put last finished first. The middle held up better: the pound and the Australian dollar finished within a place of where they were ranked. In September the same test had six of the eight finishing exactly where they were ranked. One month does not make a model, in either direction.
Month against month
| September | October | |
|---|---|---|
| Mornings measured | 21 | 19 |
| Top 3 setups the expected way | 35 of 63 | 28 of 56 |
| Top 3 together | +0.28% | -4.24% |
| Setups with the mark | 20 of 31, +3.75% | 15 of 29, -3.32% |
| First ranked beat last ranked | 12 of 21 mornings | 11 of 19 mornings |
October was worse on every line. September ran on the old model for its first two weeks; October is the first full month of the current one, so this is the month that counts against it, not for it.
November
Two central banks decide in the first week: the RBA on 3 November and the Bank of England on 5 November. The other six do not meet again until December: the Fed, the Bank of Canada and the RBNZ on 9 December, the SNB on the 10th, the ECB on the 17th and the Bank of Japan on the 18th.
So most of November is data. Every date and time is on the calendar, and what the eight look like this morning is on the strongest currency today.