← Back to Pon's Port

Kiwi&Roo

When they drift apart, add and wait.
When they come back, ride it out.

The Australian and New Zealand dollars are close neighbors that drift apart and come back together. Kiwi & Roo is an automated EA built only for AUDNZD that targets that back-and-forth with step-by-step averaging and a trailing stop.

+4,294Net profit (USD)
10,000 → 14,294
1.41Profit
factor
11.1%Max DD (balance)
(12.2% equity-based)
≈ $1,220 (≈ ¥183,000)
4,439Trades
2017.01–2026.10

* Amounts are estimates: the max DD % (equity-based) applied to the 10,000 USD starting balance (1 USD = 150 JPY).

I'm Roo, and that's Kiwi on the island next door. These numbers come from tests on past data. They're not a promise about future results, so please keep that in mind.

Backtest results from the MT5 Strategy Tester (99% history quality, $10,000 starting balance, 1:100 leverage, fixed 0.01 starting lot). They do not guarantee future results.

How Kiwi & Roo works

Planned averaging: spacing, lot size and limits are all decided in advance.

It only does two things. When price moves against it, it adds at a set distance. When price comes back, it follows as far as the move goes.

Buy 1Buy 2Buy 3Buy 4 Average price Far from the first entry Close to the average. Break-even here
1. Averaging shortens the way backAdding in steps lowers the average price, so break-even is reached without price returning to the first entry (buy-side example).
Average price (±0) Trailing starts at the threshold Peak Closed on the pullback Exit line (follows upward)
2. Trailing means no fixed profit targetOnce profit from the average price reaches the threshold, trailing begins. The exit line follows the profit upward and closes all positions together on the pullback.

Spacing widens with each step

The deeper the move against you, the wider the next spacing. It doesn't rush to add deep into a move.

Buys and sells run separately

Buy and sell baskets are managed independently, each with its own settings. A move against one side doesn't stop the other.

All positions close together

The decision is based on P/L from the average price. The deepest add and the very first position close at the same time.

Stop-out tied to balance

When the floating loss reaches a set share of the balance, everything is closed. The exit rule is fixed from the start.

Most baskets finish within a few steps

Breakdown of the baskets that closed in profit in the backtest (about 3,050).

91%Closed at step 1–2
97%Closed within 4 steps
~10 hoursTime to close
(median)
67%Closed within 24 hours

Defense built for offense

What matters most in an averaging EA is that the exit rules are set from the start.

ItemDetails
Stop-outsClose everything when the floating loss reaches a set share of the balance
Max positionsCaps the number of steps so lot size can't keep growing
Spread filterWhile spreads are wide, new entries, averaging and stop-out checks are paused
Margin checkChecks free margin before averaging and skips if short
Day-rollover guardPauses decisions for a set time after 0:00 server time to avoid misreading quotes
Retry throttle off-hoursIf closing fails because the market is closed, retries pause briefly

Backtest: about 10 years of market data

Continuous test from January 2017 to October 2026 with a $10,000 starting balance.

I won't only tell you the good parts. Only one stop-out in 10 years, but it was a big one. Please read the yellow box below too.

10,00011,00012,00013,00014,00015,000 Stop-out, Oct 2025 10,000 14,294 20172019202120232025
Balance over timeSmall wins added up, and the October 2025 stop-out cost about $1,680.
2017 +$1,1682018 +$6342019 +$7062020 +$5102021 +$2862022 +$7952023 +$5932024 +$3662025 −$1,3502026 +$586 +1,168+634+706+510+286+795+593+366−1,350+586 2017201820192020202120222023202420252026 USD. 2026 is through early October
Profit by year2025 is the only losing year, but it erased about two years of profit.

How to read these numbers (including the weak points)

  • Only one stop-out in about 10 years (October 2025, a sell basket held for a long time). It cost about $1,680, and 2025 ended at −$1,350.
  • Because there has been only one stop-out, its depth will strongly shape future results. A test with a shallower stop-out line did much worse. The current value leaves some room beyond that edge, but if a similar move goes deeper, it will be hit too.
  • This pair carries heavy swap and commission costs. In the test, about $6,640 of trading profit was reduced by about −$1,930 in swaps and about −$420 in commissions. Results depend on your broker's conditions.
  • The sell-side averaging distance setting performed noticeably better than the values around it. It may fit past markets too closely.
  • The trailing settings are the same as the EURGBP version and have not been tuned separately for AUDNZD.
  • Positions can stay open a long time. The longest single position in the test lasted about 75 days.

Who it suits, and who it doesn't

Please check whether it fits you before signing up.

Good fit

  • You can run it long-term with money you don't need
  • You accept holding a floating loss for months at a time
  • You want to watch it on a demo account or minimum lot during the free period

Not a good fit

  • You expect a profit every month or every year
  • You plan to use money you need for living
  • You want to grow your money a lot in a short time

Pricing

Both plans are free for the first 30 days. One subscription covers one account.

Start with 30 days on a demo account or the minimum lot. If it's not for you, it's fine to stop while it's still free.

Monthly plan

1,500yen / month (tax incl.)
  • 30 days free
  • MT4 & MT5
  • Use on one account
  • Cancel anytime
Start monthly plan
Save ¥6,000

Yearly plan

12,000yen / year (tax incl.)
  • 30 days free
  • ¥1,000 per month
  • MT4 & MT5
  • Use on one account
Start yearly plan

Requirements and settings

ItemDetails
PlatformMetaTrader 4 / MetaTrader 5 (MT5 requires a hedging account)
Pair / timeframeAUDNZD/M5
ParametersWorks with the default values (details of each parameter are not published)
LotsSet your starting lot (from 0.01)
Recommended setupA PC or VPS that runs 24 hours. One chart
Test conditionsMT5, 99% history quality, $10,000 starting balance, 1:100 leverage, 2017.01.01–2026.10.07

How to get started

Three steps from sign-up to running.

  1. Choose a plan and sign up

    On the payment page, enter the MT4/MT5 account number that will run the EA.

  2. Download and install the EA

    Setup guidepage: download the MT4 or MT5 version.

  3. Attach it to an AUDNZD M5 chart

    If you're logged in with the registered account, it authenticates automatically and starts running.

FAQ

Isn't averaging / martingale dangerous?

If the move against you continues, the floating loss grows. Kiwi & Roo limits that risk with a step cap, a balance-linked stop-out and a spread filter. Even so, a stop-out costs about $1,680 (backtest, 0.01 starting lot).

How is it different from Dover (EURGBP)?

Same engine, different pair and settings. Kiwi & Roo uses wider sell spacing and a deeper 11% stop-out line. It stops out less often, but each loss is larger.

Can I run Dover and Kiwi & Roo on the same account?

Yes, but when floating losses grow on both at once, margin use and losses add up. Leave plenty of margin and set different magic numbers.

Can I raise the lot size?

The test uses a fixed 0.01 starting lot. A larger lot scales the stop-out loss just as it scales profit. Decide how much loss you can accept first.

Will live trading differ from the backtest?

It can. Real spreads, swaps, execution delays and slippage change the results, and AUDNZD tends to have wider spreads. We recommend checking on a demo account or the minimum lot first.

What happens when the free period ends?

After the 30-day free period, the plan you chose is charged automatically. Cancel during the free period and you won't be charged.

How do I cancel?

The subscription management pagelets you cancel at any time. After you cancel, the EA stops opening new trades once the paid period ends.

Can I change my account number?

Yes. Each subscription covers one account, so to change it, send your new account number to .

Please read before using
  • The figures shown are backtest results on historical data. They do not guarantee future profit.
  • Averaging (martingale-style) strategies can suffer large losses if the market keeps moving in one direction.
  • Live results differ from backtests because of spreads, swaps, execution delays, slippage and similar factors.
  • Margin FX trading can lose more than your principal, and losses can exceed the margin you deposit. Trade at your own judgment and risk, with money you can afford to lose.

Take a look at the other ships too. You can compare them on the top page.