Weekly Report: The dollar rose and the euro rose with it — three FX divergences, one should converge

DXY added 0.41% last week. EURUSD also added 0.41%. Those two numbers should not coexist: the 60-day correlation between the pair and the index is -0.73, which means the euro explains most of what the dollar index does — in the opposite direction. Around that first anomaly sit two more: the RBA hiked and AUDUSD fell, and the BOJ moved a December hike to "practically assured" while USDJPY barely budged.

The week's conclusion in one sentence: the dollar's bid is broader than the Fed story, and only the EURUSD–DXY divergence is mechanically forced to close.

The arithmetic is uncomfortable. With the euro at 57.6% of the DXY basket, a 0.41% euro gain should have dragged the index down by roughly 0.24% before any other currency moved. Instead the index rose 0.41%, which means the non-euro components — the yen, the Australian dollar, the pound — sold off hard enough to more than offset the euro.

That is not a dollar story. It is a funding-currency story: the currencies being sold are the ones tied to commodities, Chinese demand and carry trades, while the euro is being held up by rate expectations that have not moved with the Fed's.

The RBA delivered its fifth hike of 2026 on Nov 3, taking the cash rate to 4.85%, and the Australian dollar fell more than 1% on the day. The reason is not the hike — it was fully priced. It is the vote: 6–3, with the statement dropping the explicit tightening language. The market read that as "December is a coin flip at best", and a hike that is delivered but not extended is a sell signal for a carry currency.

Layer on the commodity channel and the move makes sense. Copper fell 3.72% on the week and iron ore stayed soft, which narrows Australia's terms of trade just as the rate differential stops widening. AUDUSD closed at 0.6931, below the 0.70 handle it has held since late October.

The Bank of Japan held at 1.25% on Oct 30 with an 8–1 vote — Hajime Takata dissented in favour of 1.5% — and explicitly acknowledged for the first time that underlying inflation has reached 2%. Markets now price about 20bp of tightening for December, and MUFG and ING both expect a move to 1.5%. USDJPY still finished the week 0.19% lower at 157.50, and the 200-day average at 158.50 remains the level to watch.

The yen's problem is arithmetic, not intent. Even after a December hike to 1.5%, the US–Japan policy gap would be roughly 275bp if the Fed moves to 4.00–4.25% in December. A 25bp convergence does not offset a gap that size, which is why the yen needs either a faster BOJ path or a slower Fed — and last week delivered neither.

The yen is the world's funding currency, the Australian dollar is a commodity carry currency, and the euro is neither. That distinction explains the week better than any single headline.

When the dollar's bid is broad but the euro holds, the market is not buying dollars for their own sake — it is selling the two currencies that finance risk positions. That is why the correlation that normally holds (DXY against EURUSD, at -0.73 over 60 days) broke, while one that normally does not (AUDUSD against USDJPY, at -0.36) started to matter. A carry unwind shows up first in the funding legs, and it reverses only when either the Fed stops hiking or the commodity leg stabilises. Last week delivered neither: the Fed's December odds rose, and copper fell 3.72%.

The practical read is that the dollar's 102 handle is a risk-appetite gauge, not a rate-differential gauge. If copper and iron ore stabilise this week, AUDUSD should find a bid before USDJPY does — the commodity leg repairs faster than the funding leg.

Rank the three by how mechanically they must close.

EURUSD–DXY is the one that closes. The euro is 57.6% of the index; a sustained euro bid and a sustained dollar bid cannot both persist for long without one of them being wrong. Either the euro's rate support fades and EURUSD follows the dollar down, or the broad dollar bid fades and DXY gives back the 102 handle. The trigger is Tuesday's US CPI: a soft core print weakens the broad dollar bid and pushes EURUSD toward 1.13, while a hot print pulls EURUSD back to 1.1150 and lets DXY test 103.20.

AUDUSD and USDJPY are slower. AUDUSD needs either a hawkish RBA surprise in December — currently a minority view — or a stabilisation in copper and iron ore. USDJPY needs the US–Japan gap to narrow from the US side, which means the December FOMC and the 2027 dots, not the BOJ.

The condition to carry into next week: if DXY closes above 103.20 before EURUSD closes below 1.1150, the divergence is resolving in the dollar's favour and the euro's rate support was the anomaly. If the reverse happens first, the broad dollar bid was the anomaly.