DRAWUP

backtested · forex · 1,025 daily bars

nzd/usd, backtested

five classic strategies run on 1,025 days of nzd/usd, with fees, split so nothing was chosen with hindsight. two of them made money in the first half, none in the second.

best rule, second half
-0.4%
holding, second half
-6.5%
rules that beat holding
2 of 4
made money, second half
0 of 5
the DRAWUP bot holding a glowing coin

nzd/usd right now

0.55490

as of 2026-11-06, from federal reserve (fred) public data. the backtest below runs to 2026-11-06.

what this is

what we did to nzd/usd, and what came out

Five strategies, 1,025 days of nzd/usd, 2022-10-03 to 2026-11-06, fees paid on every trade. We split the history at 2025-08-15 and judged all five on the half after it, which none of them were tuned on. How the test is built.

buy and hold led the first half at +3.6%. In the half it had not seen it returned -6.5%, and rsi(14) mean reversion led instead at -0.4%. Picking on the first number would have got you the second.

Holding nzd/usd without trading returned -6.5% over that window. two of the four active rules beat it. None of the five ended up ahead, so beating the benchmark here meant losing more slowly. The gap between the best and worst rule was 8.1%, on the same prices over the same days.

the whole table

every strategy, not the winner

nzd/usd daily, 2022-10-03 to 2026-11-06. fees 5 bps a side. in sample to 2025-08-15.
strategyin sample out of sampleequity, out of sample
returnmax ddsharpe returnmax ddsharpe
rsi(14) mean reversion 3.1%-8.8%0.35 -0.4%-5.0%0.06
50/200 moving average -15.7%-16.4%-1.62 -6.6%-9.2%-0.97
20 day breakout -9.7%-18.2%-0.42 -8.5%-8.6%-1.41
12 month momentum -10.2%-11.5%-0.98 -6.0%-10.3%-1.05
buy and hold 3.6%-14.8%0.20 -6.5%-8.4%-0.72

the in sample winner was buy and hold at 3.6%. picking it on that evidence then cost 6.5% in the window that followed. that is the finding, and it is why this page prints all five.

the five

what each strategy tries to do, and what it did to nzd/usd

buy and hold — buy it on day one and never touch it again. +3.6% first half, -6.5% second, then gave it back. Worst fall 8.4%, one trade.

50/200 moving average — hold only while the recent average price sits above the long-run one. -15.7% first half, -6.6% second, and kept losing. Worst fall 9.2%, 6 trades.

12 month momentum — hold only what has already risen over the past year. -10.2% first half, -6.0% second, and kept losing. Worst fall 10.3%, 18 trades.

rsi(14) mean reversion — buy after a sharp fall, betting it bounces. +3.1% first half, -0.4% second, then gave it back. Worst fall 5.0%, 10 trades.

20 day breakout — buy when the price breaks above its highest point in twenty days. -9.7% first half, -8.5% second, and kept losing. Worst fall 8.6%, 13 trades.

the catch

where each one went wrong on nzd/usd

buy and hold — You take the whole fall with it. It is the sharpest reversal on this page: +3.6% in the half we used to form an opinion, -6.5% in the half we did not. Anyone picking on the first number got the second. It made one decision in that window. A result built on that few calls is a coin toss with a long wait attached, not an edge.

50/200 moving average — Averages only turn after the move has happened. It lost 6.6% out of sample, and -15.7% was the best it ever showed.

12 month momentum — It is buying what is already expensive. It lost 6.0% out of sample, and -10.2% was the best it ever showed. It held nzd/usd only 31.3% of the time, and that is why it lost less than holding did. Being out of the market is not the same as being right about it.

rsi(14) mean reversion — A thing that fell hard can keep falling. It still lost the least of the five out of sample, at -0.4%, having made +3.1% before that. Least-bad is the whole prize here. It held nzd/usd only 34.9% of the time, and that is why it lost less than holding did. Being out of the market is not the same as being right about it.

20 day breakout — Most breakouts fail and hand the move straight back. It lost 8.5% out of sample, and -9.7% was the best it ever showed. It held nzd/usd just 44.0% of the time and still lost 8.5%, so it was out for the recoveries as well as the falls.

year by year

which years each rule actually won on nzd/usd

calendar year returns, nzd/usd, each rule run on the same days.
yearbuy and hold50/200 moving average12 month momentumrsi(14) mean reversion20 day breakout
2022-1.3%-1.3%+0.0%+0.0%-2.1%
2023+1.3%-2.6%+1.9%+1.7%+0.8%
2024-0.3%-1.6%-5.4%+0.2%-3.3%
2025-5.4%-0.8%-2.6%-2.4%-3.2%

so what works

the hard part is not the trading

Nothing worked on nzd/usd in the half it had not seen. Running a rule without flinching is the easy part for a machine; knowing which rule is worth running is not, which is why we test this many and publish the failures too.

reading it

what this says

nzd/usd fell through the out of sample window and not one of the five made money. the honest reading is not that these five are bad, it is that a backtest chosen on past performance tells you little about the next window. the federal reserve publishes this series on a schedule rather than continuously, so the window ends on its last published day, 2026-11-06.

the one that lost least, rsi(14) mean reversion, held for 35% of the window. sitting out is the only edge visible here, and fees and funding could erase it.

method

how it was run

data
nzd/usd daily closes, federal reserve (fred) public data, 1,025 bars, 2022-10-03 to 2026-11-06, read 2026-11-16. this series is published as one value a day with no intraday range, so the 20 day breakout is measured on closing prices. holding one currency against another also earns or pays the gap between their two countries' interest rates, which a price series does not carry, so none of these figures include it.
fees
5 basis points a side on every change of position. carry is not modelled. holding this pair earns the interest-rate gap when the base currency's rate is the higher one and pays it when it is not, so the omission does not run one way. it is largest where the gap is widest.
no hindsight
every decision is taken on one close and applied to the next day's move.
the split
first 70% sets expectations, last 30% tests them. nothing was tuned on it.
missing
one asset, one timeframe, long only, no sizing, no stops.

this market

what stands out about nzd/usd

Held and not traded, nzd/usd returned -6.5% out of sample, 38th of 144 markets measured the same way.

It was unusually calm for this set. The worst fall while holding was 8.4%, shallower than 135 of the other 143 markets.

2 of the four rules beat holding. The best, rsi(14) mean reversion, returned -0.4% out of sample against -6.5% for doing nothing.

That rule made +3.1% in the first half and -0.4% in the second. A rule that changes sign between halves has not found anything in nzd/usd; it fitted the first half.

The gain was unusually evenly spread: the five best days account for only 9.0% of it, so the result does not rest on catching a handful of sessions.

rsi(14) mean reversion was only in the market 34.9% of the time and traded 10 times across 1,025 days, so most of that return was earned while holding nothing.

The best rule never falls 10% below its own high water mark, so it would have survived a funded account's trailing drawdown for the whole period. Few of the 144 markets manage that.

Year by year, holding nzd/usd gave 2022 -1.3%, 2023 +1.3%, 2024 -0.3%, 2025 -5.4%. That is 1 up years and 3 down.

The four rules, best to worst out of sample on nzd/usd: rsi(14) mean reversion -0.4%, in the market 34.9% of the time, worst fall 5.0%; 12 month momentum -6.0%, in the market 31.3% of the time, worst fall 10.3%; 50/200 moving average -6.6%, in the market 48.5% of the time, worst fall 9.2%; 20 day breakout -8.5%, in the market 44.0% of the time, worst fall 8.6%.

questions

what people ask

what is the best strategy for nzd/usd?

rsi(14) mean reversion came out best out of sample at -0.4%, against -6.5% for holding and not trading. 2 of the four rules beat holding.

did the winning rule keep working on nzd/usd?

no. rsi(14) mean reversion made +3.1% in the first half and -0.4% in the second. a rule that changes sign between halves fitted the first half rather than finding anything in nzd/usd.

would nzd/usd have passed a funded account challenge?

on these rules it survives the usual limit. rsi(14) mean reversion never falls 10% below its own high water mark across the whole out of sample window, which few of the 144 markets we measure manage.

how often was the best rule actually holding nzd/usd?

34.9% of the time, across 10 trades in 1,025 days. most of the period it held nothing, which is why its worst fall is shallower than the market's.

which years were good for nzd/usd?

2022 -1.3%, 2023 +1.3%, 2024 -0.3%, 2025 -5.4%. 1 up and 3 down.

bots we have actually tested, with the results published

more

other backtests