EUR/GBP and USD/JPY moved almost independently into September 2026

Over the 40 ECB fixings to 28 September 2026, the daily changes in EUR/GBP and USD/JPY had a correlation of -0.13. That is our own calculation from European Central Bank reference rates, and it points to almost no link between the two pairs over that stretch.

A reading that close to zero means knowing what EUR/GBP did on a given day told you next to nothing about what USD/JPY did on the same day. The two were not offsetting each other in any reliable way, and they were not tracking each other either. They were simply moving on their own separate causes.

Why the window is set at 40 fixings

Our correlation figures use the daily changes of the previous 40 ECB fixings. A shorter window can shift sharply on a single outlier day, so 40 fixings gives a steadier read on whether two pairs are leaning on each other or not. For this pairing we have calculated the figure on that window only, to 28 September 2026.

A correlation of -0.13 over the 40 ECB fixings to 28 September 2026 means daily changes in EUR/GBP and USD/JPY gave almost no information about each other. On most days, knowing which way one pair moved said nothing about the other.

Why that changes the arithmetic for a combined book

Where two positions move together, the swings in one tend to add to the swings in the other, for better or worse. Where they move independently, the gains and losses on one side have tended to land on different days from the gains and losses on the other, so the combined daily swing over this period came out smaller than the two added together. That is the mechanical effect of a correlation near zero.

What the figure does not say

It says nothing about why the two pairs decoupled over this particular stretch, and it carries no view on whether that holds into October. Bank of England and Bank of Japan policy run on separate timetables and separate drivers, which is one plausible mechanism for two pairs trading without reference to each other, but the correlation figure itself does not distinguish mechanism from coincidence.

It is also worth setting beside the same 40-fixing window for EUR/USD and USD/CHF, which moved in opposite directions on almost every day the coefficient was calculated over. The contrast shows that near-zero correlation is not the default state for a currency pair. Some pairs move in step, some move in near mirror image, and some, like this one, move apart.

Reading the number for what it is

-0.13 is a correlation of daily changes over a stated 40-fixing window ending 28 September 2026, calculated from European Central Bank reference rates. It is our own calculation, it could contain errors, and it describes a window that has already closed by the time it is read. A different 40 fixings, earlier or later, would very likely return a different number.

Sources

The correlation figure behind this analysis is our calculation from ECB reference rates, the euro foreign exchange rates the European Central Bank sets at approximately 14:10 CET and publishes for information purposes only, not as dealing prices. The window it covers and the strength label attached to it are also our own calculation; neither the ECB nor the Bank of England nor the Bank of Japan publishes a correlation figure for this pairing.

Background on the central-bank decisions and economic releases that move sterling and the yen over the same stretch is kept on the central banks page and the economic data page, each dated and sourced to the relevant official publisher.

What would change our view

A correlation measured over 40 ECB fixings describes a period that has already closed. It says nothing about whether EUR/GBP and USD/JPY keep moving independently once a driver appears that touches both at once, such as a shift in risk appetite broad enough to pull a funding currency and a safe haven the same way regardless of the UK or Japan's own policy path. If that happened, running the same 40-fixing window afresh over the following weeks would show the correlation moving away from zero.

A shift in central-bank forward guidance from the Bank of England or the Bank of Japan, landing close enough in time to move both pairs within the same stretch of fixings, is the most likely single event to alter this reading. Scheduled decisions and data releases that could do that are listed, with their date, time and source, on the events calendar.

Corrections

No corrections have been made to this piece since it was published. Any future correction will be dated here and will state what changed and why.