AUD/USD printed 0.66190 on the last tick of the week, with the platform feed showing a bid of 0.66180 and an ask of 0.66200 — a one-pip spread in a week when nothing about this pair was calm. The high was 0.66840, the low 0.66124, and the pair closed 0.62% lower, its third consecutive weekly decline.
The interesting part is not the number. It is the divergence behind it. The Reserve Bank of Australia raised the cash rate to 4.85% on 3 November, the highest since late 2008, and the currency is near an eleven-month low. Both facts are true at once because AUD/USD does not price Australian policy. It prices the difference between Australian and American policy, and that difference is shrinking.
Research note: prices in this article are the Friday 20 November 2026 platform prints (21:59-22:30 UTC), verified against the feed at the time of writing. Policy facts are taken from the RBA’s own media release 2026-30 and the ABS Labour Force release for October 2026. Nothing here is a forecast of a level.
The 3 November decision, because the vote matters more than the hike
The Board lifted the cash rate target by 25 basis points to 4.85%, from 4.60%. It was the fifth increase of 2026, taking the total tightening this year to 1.25 percentage points. It was not unanimous: six members voted for the increase and three voted to hold at 4.60%.
That split is the story. Six-to-three is the closest this Board has come to a genuine argument all year, and the dissenters were not arguing about inflation being contained — they were arguing that the transmission is already working. Their case, as reported from the minutes published on 17 November, is that housing prices have fallen significantly, the unemployment rate has risen, and further tightening risks a sharper downturn than the Board intends.
The majority case is in the statement itself: inflation is broadly based, the staff forecasts have been revised materially higher, and inflation is now not expected to reach the top of the 2-3% target range until late 2027, roughly three quarters later than the previous forecast. The Board also named the drivers, and they are worth repeating because they are not the ones retail traders assume: higher fuel and commodity prices from the Middle East conflict, domestic capacity pressures judged tighter than previously thought, and AI-related investment driving rapid growth in prices for technology-related goods.
What the statement does not contain is guidance. There is no line promising further increases if needed, and when Governor Michele Bullock was asked what comes next she said, plainly, “I don’t know what the next move will be”. She also noted that the last two hikes “haven’t even started to flow through yet”. A central bank that says both of those things in the same press conference is a central bank that has stopped pre-committing.
Then the labour market broke the forecast
The November Statement on Monetary Policy, with forecasts finalised on 28 October, projected unemployment at 4.7% for December 2026, rising to 5.1% by end-2028. GDP growth of 1.6%, CPI of 4.5% for December and a trimmed mean of 3.6%.
On 19 November at 11:30am AEDT, the ABS reported October labour force data:
| Metric | September 2026 | October 2026 | Change |
|---|---|---|---|
| Unemployment rate | 4.7% | 4.9% | +0.1 pts (highest since October 2021) |
| Employed people | 14,782,200 | 14,738,600 | -43,500 (-0.3%) |
| Unemployed people | 735,800 | 753,700 | +17,800 (+2.4%) |
| Full-time employment | — | 10,228,700 | -2,900 |
| Part-time employment | — | 4,509,900 | -40,600 |
| Participation-side ratio (emp/pop) | 63.6% | 63.3% | -0.3 pts |
Read the composition carefully, because it is the difference between a soft patch and a warning. Part-time employment fell by 40,600 while full-time fell by only 2,900. Households are not being laid off en masse; the hours are being cut at the margin, and the unemployment rate rose because 17,800 more people were counted as unemployed rather than because the labour force expanded.
Markets did the arithmetic immediately. The odds of a December hike collapsed, the Australian dollar fell about half a cent, and the conversation moved from “how high does the RBA go” to “when does it cut”. Reuters reporting after the print had traders pricing roughly 45 basis points of cuts by end-2027.
One month does not settle a quarterly average, and the December forecast is an average, not a single print. But a forecast that is already 0.2 points behind after the first month of the quarter is a forecast the Board will have to address in the statement, not the minutes.
Where the pair actually gets its direction: the Fed leg
AUD/USD is a two-legged trade and the second leg is doing most of the work. The Federal Reserve has held the funds rate at 3.75-4.00% since its September hike, held again on 28 October on a 10-2 vote with two dissents arguing for a hike, and futures price roughly 90% odds of another 25bp increase at the 8-9 December FOMC.
So the market’s expected gap is moving the wrong way for the Aussie. Australia at 4.85% and holding, the US at 3.75-4.00% and hiking: the compression comes from the top. Add a global deleveraging week — the dollar index printed 103.840 with a high of 103.953, gold fell 3.98% to 3,539 and equity indices sold off hard — and the currency does what a liquid, high-beta commodity currency always does in that tape. It gets sold, and it gets sold first.
The carry arithmetic, honestly
Here is the part that sounds like free money and is not. With the RBA at 4.85% and the funds rate at 3.75-4.00%, a long AUD/USD position is paid to be held, roughly 0.85 to 1.10 percentage points annualised on the interest differential. That is a genuine change: for most of this cycle the Aussie was the funding currency, and shorts collected.
Work it at a realistic size. On 0.01 lots of AUD/USD at 0.6619, notional is about A$1,000, or roughly US$662. At an annualised differential of about 1%, the daily financing credit is a few cents, and the spread you pay to enter is one pip — about US$0.10 here. A month of positive carry is therefore wiped out by an adverse move of roughly 0.08%, which is less than a tenth of Friday’s range.
That ratio is the whole argument against trading this pair for carry. The carry is real, it is simply smaller than the noise floor of the instrument. It becomes interesting only at size, held for quarters, with the option of not being stopped out — which is a different trade from a leveraged CFD position with a defined stop.
Level map, with roles rather than promises
Friday’s range gives the frame: high 0.66840, low 0.66124, close 0.66190.
- 0.6604 — the 78.6% retracement of the 2026 range and the first level that has to hold. It sits below Friday’s low, which means it has not been tested yet; that is exactly why it matters.
- 0.6420 — the next shelf if 0.6604 gives way on a weekly close.
- 0.6748 — the 61.8% retracement, first resistance on any recovery.
- 0.6800 — round number, and the level where the market will look for sellers.
- 0.6838 — the 20-day EMA, which has capped every rally attempt since the pair broke down.
None of these are floors. A stop placed exactly at a published level is a donation to whoever is running the liquidity above it; place it beyond the level and size the position from the stop, not the other way round.
Three scenarios into 8 December
The Board meets on 7-8 December and announces on the 8th, the last meeting of the year.
Hold at 4.85% with no guidance (roughly the base case). This is the outcome the labour market data now favours and the one that most hurts the Aussie, because it confirms the hiking cycle is over without offering a cut. Expect 0.6604 to be tested within days; a break opens 0.6420 and the pair spends December drifting rather than trending.
Hold, but with an explicit easing bias. The three dissenters get their way in the language even if not in the vote. This is the single most bullish scenario for AUD/USD — not because a cut is coming soon, but because the market would re-price the whole 2027 path. A move back through 0.6748 would be the first evidence.
Hike to 5.10% on a re-acceleration in the 25 November CPI. Now a minority scenario after the employment print, but not a negligible one: the Board has already said inflation will not reach the top of the band until late 2027. A hike here would be the sixth of the cycle and would put the Aussie at an 18-year-rate differential that finally compensates for the risk — with the caveat that a hiking central bank into rising unemployment usually rallies the currency for a week and then loses it.
What actually matters before the decision
October monthly CPI, 25 November — the last inflation print before the Board meets, and the one number that can revive the hike scenario. Then Q3 capex on 26 November, the day after US Thanksgiving, when liquidity thins and ranges widen for reasons that have nothing to do with Australia.
The terms of trade is the other leg nobody watches. Iron ore and the commodity complex are the only channel that can rescue the Aussie if the rate channel keeps pointing down, and the RBA’s own statement names commodity prices as a driver of the revised inflation path. If the Middle East conflict de-escalates, both the inflation problem and the terms-of-trade problem resolve in the same direction.
Trading it, and what the platform changes
On UZFX the relevant instruments are live and quotable in one account: 26 currency pairs including AUD/USD and AUD/JPY, four metals, three energy CFDs, three crypto CFDs, seven index CFDs and three stock CFDs — 100+ instruments in total, with a $10 minimum deposit, a 0.01-lot minimum and spread-only pricing on standard accounts. There is no ECN tier and no commission line, which matters for a pair where you are counting pips: a 1.0-pip spread at 0.66190 is the entire cost of entry.
There is also no MetaTrader 4 or MetaTrader 5 — execution runs through the Web Terminal, H5 mobile, and the iOS, Android, Windows and Mac apps. For a discretionary trade around a central bank decision that is not a limitation; the platform carries the same depth on AUD/USD and AUD/JPY that any desktop terminal would show. What it does change is the discipline requirement: a one-click mobile ticket with no separate risk module makes it easy to size by feel. Size by stop instead.
Two structural facts worth stating plainly, because they are the ones that decide whether a bad week becomes a bad year. UZFX is regulated by ASIC under AFSL 001291473 and applies negative balance protection, so a gap through a stop cannot take an account below zero. Retail leverage on major pairs under ASIC rules is capped at 30:1, which on a 0.01-lot AUD/USD position means roughly US$22 of margin rather than the US$1.30 an offshore 500:1 entity would ask for. That cap is not an inconvenience to be worked around. It is the only thing standing between an 18-year-high rate differential and a margin call on a 40-pip move.
FAQ
Q: Why is AUD/USD falling while the RBA is raising rates?
Because the pair trades the gap between Australian and US rates, not the level of either. The RBA is at 4.85% and the Federal Reserve is at 3.75-4.00%, but the market expects the Fed to hike again in December and expects the RBA to stop, so the expected gap is narrowing rather than widening. A global deleveraging week does the rest: the Australian dollar is a liquid risk proxy and gets sold first.
Q: What is the RBA’s next decision date?
The Board meets on 7-8 December 2026 and announces the decision on 8 December, the last meeting of the year. The October monthly CPI indicator is published on 25 November, one day before US Thanksgiving thins liquidity.
Q: Does a long AUD/USD position earn positive carry at 4.85%?
Yes, for as long as the RBA holds and the funds rate stays at 3.75-4.00%. On a 0.01-lot position the annualised carry is worth roughly 0.85 to 1.10 percentage points, but a single 0.8% adverse move wipes out about a year of it.
Q: What level matters most on the downside?
0.6604, the 78.6% retracement of the 2026 range and the first real support under Friday’s low of 0.66124. A weekly close below it opens 0.6420. None of these are floors, and a stop placed at the level itself will be filled by noise.
Q: How do you check that a broker is actually licensed before trading an event week?
Search the regulator’s own register by legal entity name rather than by brand, confirm the licence is active, and confirm the licensed activity covers dealing rather than marketing. The ASIC professional registers search is the starting point in Australia; the step-by-step register walkthrough covers the FCA, CySEC and NFA equivalents.
Related reading
- AUD/USD at 0.66: the RBA’s 4.60% hike and why the pair kept falling — the prior chapter, written before the November decision
- RBA August 2026 decision: AUD/USD forecast and strategy
- AUD/JPY: the Asian cross-pair strategy guide — for the risk-sentiment leg of the same trade
- DXY at 103.8: dollar strength pressuring Asian FX — the frame for the Fed leg
- USD/JPY and the carry-trade unwind — the other side of the same funding trade
Risk warning
Leveraged forex and CFD trading carries a high risk of loss and can lose more than the amount deposited, even with negative balance protection on gaps that occur outside guaranteed protection. This article is market analysis and education, not investment advice, and it makes no prediction about the level of AUD/USD or the RBA’s next decision. Rates, spreads and leverage terms differ by the entity that holds your account; verify them in your own client agreement before sizing a position.
Last reviewed: 21 November 2026. Editorial team, MarketCFD.com.