In this article
After this article, you can
- Separate a predetermined range from hindsight fitting.
- Count observations with an explicit endpoint rule.
Terms at hand: Backtesting · Trading Plan · Trading Journal · Position notional (notional value) · Balance · Liquidity
AUD/NZD compares two currencies exposed to commodity exports and regional demand. Some shared influences may offset when both currencies move similarly, but global risk does not disappear. Relative RBA and RBNZ policy expectations, trade conditions and positioning can all affect the cross.
Two antipodean cousins in numbers
AUD/NZD quotes New Zealand dollars per Australian dollar: 1.0800 means NZD 1.08 per AUD. The conventional pip is 0.0001 NZD per AUD. For a hypothetical 10,000 AUD position, a one-pip change represents NZD 1 before costs and any conversion to the account currency. A range visible over one period is not a permanent feature of the pair.
Shared exposure can moderate a cross-rate move when both currencies respond similarly. Their responses need not be identical, so AUD/NZD can still react to changes in global risk appetite.
A fitted range does not repair an earlier test
A range can describe a chart segment well. Separate a rule fixed beforehand from bounds widened only after seeing new observations.
Scroll sideways to see the full diagram
Assumptions: Our fictional AUD/NZD reference rates at ten equal intervals. The range [1.08; 1.12], endpoint inclusion, observation order and no-change rule are fixed before assessment. This is a constructed example, not market data, Bid/Ask, trades or proof of pair stability. No intervening path is assumed. The wider range is a later description only.
Before reading, record 1.08 ≤ rate ≤ 1.12 and assess exactly the next ten readings without changing bounds. 1.08 and 1.12 are inside because endpoints are included. Reading 6 at 1.06 and reading 8 at 1.14 fail the criterion. Original result: 8 / 10 = 80%.
After seeing the data, describe it with 1.06–1.14. But 10 / 10 answers a different question. Do not replace the earlier result with it or call it a prediction.
MetaQuotes — separate fitting and checking periods describes checking selected parameters on a separate period. Our diagram is a simple membership test, not a strategy backtest. For a rule change, record a new version and another unseen set. Ten points establish neither financial returns nor the next move.
Cross mechanics: when shared moves offset
At consistent reference rates from the same instant, AUD/NZD = AUD/USD ÷ NZD/USD. USD per AUD divided by USD per NZD gives NZD per AUD. In a hypothetical example without spreads or fees, 0.6600 ÷ 0.6000 = 1.1000. If both dollar pairs rise by 1%, 0.6666 ÷ 0.6060 remains 1.1000. If AUD/USD rises by 2% and NZD/USD by 1%, 0.6732 ÷ 0.6060 is about 1.11089: a 0.9901% cross increase, not exactly 1%. The exact multiplier is 1.02 ÷ 1.01; subtracting the two percentage changes is an approximation. Mid-rate arithmetic does not establish an executable Bid/Ask quote or an arbitrage opportunity.
Equal units on two legs do not cancel USD cash flows. In a hypothetical cash exchange without fees, buying 10,000 AUD at AUD/USD 0.6600 costs USD 6,600. Selling 10,000 NZD at NZD/USD 0.6000 raises USD 6,000, leaving a USD 600 funding requirement. Selling 11,000 NZD raises USD 6,600 and balances those USD flows. This cash example is not a universal CFD hedge formula: derivative contract sizes, settlement, margin, execution timing and financing also matter.
What really drives it
Relative monetary-policy expectations matter for AUD/NZD. A more hawkish surprise from one central bank may support its currency, but the reaction also depends on prior pricing and other market conditions.
Terms of trade compares export prices with import prices. RBA describes commodity export prices as one influence on AUD. In its February 2026 statement, RBNZ discussed dairy, meat and horticulture export prices and attributed the historical NZD trade-weighted appreciation to higher domestic interest rates and a weaker USD. This basket observation neither establishes a current AUD/NZD forecast nor a rule that dairy outperforming metals must weaken the cross. Read the separate contexts in the AUD/USD pair and the NZD/USD pair articles.
Who it suits and which style
A mean-reversion strategy requires a defined sample, fixed range rules and testing that includes trends and transaction costs. A range visible in hindsight does not establish a reliable trading edge or a predictable future rhythm.
Fix the observed period, range bounds and invalidation rule before evaluating a range hypothesis. Moving the bounds after a breakout changes the test. Diverging policy expectations are one possible influence; a rate differential alone does not show that mean reversion will succeed or identify the point at which it fails. For example, a hypothetical policy-rate rise from 4.00% to 4.25% is 0.25 percentage points, or 25 basis points, and a 6.25% relative increase in the rate itself. None is an exchange-rate return.
What to do before you open an AUD/NZD position
Monitor Sydney and Wellington trading activity and both Australian and New Zealand release calendars. Compare spreads and execution for your actual provider and time window. Convert event times to your own zone, then choose a schedule compatible with your trading rules; your location alone does not determine a suitable strategy.
Budget for a breakout and a worse fill than an ordinary stop level; a stop does not guarantee the intended exit price. Retail overnight financing is also separate from the RBA–RBNZ policy-rate gap. OANDA Europe, for example, describes FX rates based on liquidity-provider tom-next rates plus its administration fee, with separate long and short rates. Check the actual product, direction, units and day multiplier; a financing credit cannot eliminate an adverse exchange-rate move.
AUD/NZD expresses the relative performance of Australia and New Zealand. Compare both policy calendars and trade exposures while retaining global risk and liquidity in the analysis.
Check your understanding
One short exercise
About 2 min · you can try again
The fixed range [1.08; 1.12] has two of ten points outside. What does later widening to [1.06; 1.14] change?
- The description criterion; the original result remains 8 / 10.
- The original result to 10 / 10 because all points can be included.
- It guarantees point 11 will be inside.
Check the answer and explanation
The description criterion; the original result remains 8 / 10.
The wider range includes the sample but was not the initial rule. It does not repair its result retrospectively.
Frequently asked
Does AUD/NZD always trade in a range?
Which matters more for the rate — the RBA or the RBNZ?
Is AUD/NZD a good pair for a beginner?
When is AUD/NZD most liquid?
Sources & bibliography
-
BIS Triennial Central Bank Survey 2022 · Official foreign-exchange turnover statistics for the 2022 survey. www.bis.org ↗
-
RBA Monetary policy · Official monetary-policy information from the Reserve Bank of Australia. www.rba.gov.au ↗
-
RBNZ Monetary policy · Official monetary-policy information from the Reserve Bank of New Zealand. www.rbnz.govt.nz ↗
-
RBA Exchange Rates and their Measurement · Bilateral quotation units and calculation of cross rates through a third currency; reference-rate arithmetic is not an executable quote. www.rba.gov.au ↗
-
RBNZ Monetary Policy Statement February 2026 · Historical explanation of NZD trade-weighted appreciation and export-price influences, not a current AUD/NZD forecast. www.rbnz.govt.nz ↗
-
OANDA Europe Financing Costs · Provider-specific FX tom-next funding, administration fees and separate long/short rates; not a simple central-bank rate difference. www.oanda.com ↗
