A big move or a normal day? Tracking the major currencies

The European Central Bank set its reference rates for 6 October 2026, and that fixing is the latest this site has published for the pairs it tracks.

Each rate below is the ECB reference rate for that date, set at approximately 14:10 CET for information purposes. The one-day change and the multiple beside it are our own calculations from those rates, and both could contain errors.

Rates, one-day changes and normal-day multiples

The normal-day multiple states each move against that pair's typical day, the standard deviation of its daily changes over the previous 20 ECB fixings. A multiple of 1.0 means the move matched that average; a multiple below 1.0 means it was quieter than usual.

EUR/USD has its own page with further detail on its range and correlations.

EUR/USD moved furthest from its typical day of any pair at the ECB fixing of 6 October 2026, rising 0.58% on the previous fixing, a move 2.2 times its typical day over the previous 20 fixings.

Between the ECB fixing of 21 September 2026 and that of 28 September, the US dollar rose against five of the six major currencies this site tracks against it. Only the yen moved the other way. These are our own calculations from European Central Bank reference rates, set at about 14:10 CET and published for information rather than for dealing; the calculations could contain errors and are already out of date by the time anyone reads them.

• The euro fell 0.97% against the dollar.
• The pound fell 0.98% against the dollar.
• The Swiss franc fell 1.25% against the dollar.
• The Canadian dollar fell 1.14% against the dollar.
• The Australian dollar fell 1.59% against the dollar, the largest fall of the six.
• The yen rose 0.25% against the dollar, the only gain of the six.

Where the move showed up on the quoted pairs

Quoted the way these pairs are traded, USD/CHF rose 1.26% over the week and stood at 0.8318 at the fixing of 28 September, its highest level since at least 1 July. USD/CAD rose 1.15% over the same stretch. USD/JPY fell 0.25%, the mirror of the yen being the one currency that gained ground against the dollar.

What a move shaped like this usually involves

A move spread across five currencies at once, with only one holding out against it, points to something on the dollar side of each pair. Whether that was demand for dollar assets, a shift in expected interest-rate paths, or positioning ahead of scheduled data is not something a single week of ECB fixings can settle by itself.

What is not yet established

There is not yet a sourced account of which central-bank statement, data release or event was mainly responsible for the week's dollar move. Decisions and releases from the Federal Reserve, the ECB, the Bank of England, the Bank of Japan and the Swiss National Bank over the same stretch are the first place to look, and are covered on this site as they are published, alongside the pairs they move.

A fuller account of the week, with its timeline and sources, is set out in The Dollar Rose Against Every Major Currency Except the Yen, 21–28 September 2026.

A central-bank decision or an inflation release can move a pair further in a few minutes than a normal trading day moves it. Knowing one is due, and which pairs it tends to affect, is part of reading any move that follows it.

The events calendar lists what is coming, each entry dated and timed from the publisher's own schedule, with the pairs it typically affects and the last published figure for that release. Where a consensus figure is shown, its source is named alongside it; where none has been published, none is shown.

Where the releases come from

Central-bank decisions are drawn from the official statements of the Federal Reserve, the ECB, the Bank of England, the Bank of Japan and the Swiss National Bank, and set out on the central banks pages alongside the pair moves that followed past decisions. Inflation, jobs and growth releases come from official statistical publishers and sit on the economic data pages with the same treatment: date, source, and the pairs each one moved.

None of this is a forecast of what a release will show or what a pair will do when it lands. It is a dated note of what is due, so a move can be read against the event that produced it.

At the ECB fixing of 6 October 2026, the eight pairs this site tracks moved by different amounts, and some of those moves were larger relative to what is normal for that pair than others. Ranking them this way puts a quiet session and a busy one on the same scale.

Typical day and normal-day multiple

A pair's typical day is the standard deviation of its daily changes over the previous 20 ECB fixings, in pips. The normal-day multiple states a single day's move as a fraction or multiple of that typical day, so a move of 2.2 means the pair moved roughly twice as far as it usually does, and a move of 0.1 means it barely moved at all. Both figures are our own calculation from European Central Bank reference rates. They could contain errors, and by the time they are read they are already out of date.

Ranked by normal-day multiple, 6 October 2026

1. EUR/USD: 2.2 times its typical day
2. GBP/USD: 1.3 times its typical day
3. EUR/JPY: 1.2 times its typical day
4. AUD/USD: 1.1 times its typical day
5. EUR/GBP: 1.0 times its typical day
6. USD/JPY: 0.2 times its typical day
7. USD/CHF: 0.1 times its typical day
8. USD/CAD: 0.1 times its typical day

EUR/USD stands out on this list, moving more than twice its typical day while USD/CHF and USD/CAD barely moved at all. A reading further down the analysis on EUR/USD sets out what drove that move and what would change that view.

Two currency pairs can drift in the same direction, move in opposite directions, or carry on with barely any relation to each other at all. Over the 40 ECB fixings (the daily reference rate set by the European Central Bank at approximately 14:10 CET) to 6 October 2026, two pairings stood out for different reasons.

• A correlation below 0.2 in size counts as almost no link.
• A correlation from 0.2 up to 0.7 counts as moderate.
• A correlation of 0.7 or above counts as strong.
• A negative figure means the two pairs moved against each other.

EUR/USD and USD/CHF

Daily changes in EUR/USD and USD/CHF carried a correlation of -0.53 over the period, a moderate opposite link. A move in one pair tended to come with a move against it in the other, so a position in both was, over this stretch, partly cancelling itself out. The EUR/USD page sets out that pair's own range and typical day in full.

EUR/GBP and USD/JPY

Daily changes in EUR/GBP and USD/JPY carried a correlation of -0.13 over the same 40 fixings, which counts as almost no link. Each pair was largely answering to its own drivers, independent of what the other was doing. EUR/GBP and USD/JPY Moved Almost Independently to September 2026 sets out the fixings behind that figure in more detail.

Both correlations are our own calculation from European Central Bank reference rates. Like any calculation run against a moving market, they can contain an error, and the window moves on with every new fixing, so a figure calculated today is already behind by the time it is read.

Each piece opens with the finding, then explains the mechanism in plain terms, then sets out the full account with sources and what would change the reading. The three most recent are below.

The dollar against every major currency but the yen

Over the week of 21 to 28 September 2026, the dollar strengthened against every other major currency this site tracks except the yen. The piece sets out which pairs moved, the data behind each move, and what would need to change for the pattern to reverse.

AUD/USD's sharpest day

AUD/USD had its largest one-day move of September 2026, measured against its own typical day rather than against the other pairs on the dashboard. The analysis traces the move to its stated source and notes where the account is still incomplete.

EUR/GBP and USD/JPY apart

EUR/GBP and USD/JPY moved with little relation to each other through to September 2026, despite both touching the dollar and the pound indirectly. The piece gives the correlation window and date range behind that reading, and states what would be needed to call it more than coincidence.