Key results
- Total return
- -68.6%
- Buy & hold: +55.9%
- Annual return (CAGR)
- -10.9%
- Buy & hold: +4.5%
- Max drawdown
- -69.8%
- Buy & hold: -15.9%
- Win rate
- 30.44%
- Profit factor
- 0.75
- Trades
- 2,474
- Longest losing streak
- 26 trades
- Period
- Oct 2, 2016 to Oct 1, 2026
- 10 years
- Market
- USD/JPY
- 1-hour
Equity curve vs buy & hold
Returns year by year
| Year | Return | Trades |
|---|---|---|
| 2016 | +1.3% | 61 |
| 2017 | -5.6% | 236 |
| 2018 | -7.7% | 203 |
| 2019 | -13.1% | 169 |
| 2020 | -12.8% | 200 |
| 2021 | -11.4% | 203 |
| 2022 | -10.1% | 300 |
| 2023 | -14.1% | 317 |
| 2024 | -9.6% | 289 |
| 2025 | -14.8% | 306 |
| 2026 | -11.3% | 190 |
Calendar years, compounded from monthly results. The first and last years can be partial.
Three white soldiers and its mirror, three black crows, are classic Japanese candlestick patterns: three strong candles in the same direction, each closing beyond the previous one. They are the most mechanical way to describe a burst of momentum, which makes them a natural candidate for a systematic test. Strategy collections such as Quantified Strategies' free strategies list include them. We added a 200 EMA trend filter and ran them on ten years of hourly USD/JPY, a pair that alternated clear up and down phases over the decade.
The rules we tested
- Market and timeframe: USD/JPY, 1-hour candles.
- Period: 2 October 2016 to 1 October 2026.
- Long entry: a three white soldiers pattern appears and the close is above EMA(200).
- Short entry: a three black crows pattern appears and the close is below EMA(200).
- Stop-loss: 30 pips. Take-profit: 60 pips, twice the risk.
- Exit: the stop, the target, or an opposite entry signal, which closes the trade and reverses the position.
- Costs and sizing: 0.02% commission per side, full position size on a 10,000 starting balance.
What the backtest shows
The result is a heavy loss. The account lost 68.6% over ten years (-10.9% a year), with a maximum drawdown of 69.8%. Holding USD/JPY returned +55.9% with a drawdown of 15.9%.
The bracket needed a win rate of about one in three to break even before costs. The strategy won 30.4%. Of 2,474 trades, 1,416 hit the 30-pip stop, 749 reached the 60-pip target and 308 closed on an opposite signal. The profit factor of 0.75 means gross losses were about a third larger than gross wins.
Frequency turns a small shortfall into a large loss. Around 250 trades a year at 0.04% per round trip add up to roughly the full position size over the decade. With a stop of only 30 pips, roughly 0.2% to 0.3% of USD/JPY's price over the decade, the 0.04% round-trip commission eats a meaningful slice of the risk on every trade.
Every full calendar year was negative. Only the partial 2016 (+1.3%) finished in profit. The worst years were 2025 (-14.8%) and 2023 (-14.1%), and even 2022, a strong trend year for the pair, lost 10.1%.
A pattern is an event, not an edge
Three strong candles describe what just happened. Whether it predicts the next move is exactly what a backtest measures, and here the answer was no.
Why it doesn't work here
Three consecutive strong candles on an hourly chart usually come at the end of a short burst, not the beginning. By the time the third candle closes, price has already travelled, short-term traders take profit, and a pullback follows. A 30-pip stop on USD/JPY sits well within the normal size of such pullbacks, which is why more than half of all trades were stopped out. The 200 EMA filter does keep trades on the side of the trend, but it cannot fix an entry that consistently arrives late in the move.
What you could test next
- Wait for a pullback after the pattern before entering, for example a return to a short EMA.
- Widen the stop and target (60 and 120 pips) so normal retracements do not close the trade.
- Move to the 4-hour or daily chart, where three strong candles represent a more significant shift.
- Test the opposite reading: treat three white soldiers above the 200 EMA as a short-term exhaustion signal and measure the fade.
Methodology and assumptions
- Starting capital
- 10,000
- Position size
- 100% of equity
- Commission
- 0.02% per side
- Data window
- Oct 2, 2016 to Oct 1, 2026
- Run date
- October 2, 2026
- No repainting: every signal is computed on closed candles only, so the backtest never acts on a price a trader could not have seen yet.
- Past performance does not predict future results. A backtest is a historical simulation, not a forecast.
Frequently asked questions
What is the three white soldiers candlestick pattern?
Three consecutive bullish candles, each closing higher than the last, read as a strong push of buying momentum. Its bearish mirror is three black crows: three consecutive bearish candles closing lower each time.
Is the three white soldiers pattern reliable?
Not as a standalone entry on hourly USD/JPY. With a 200 EMA trend filter, a 30-pip stop and a 60-pip target, it won 30% of 2,474 trades, below the roughly one in three a 1:2 bracket needs to break even.
Why does a trend filter not help the three white soldiers pattern?
The 200 EMA keeps trades on the side of the trend, but three strong candles in a row often mark a short-term exhaustion point. Entering after the push means entering when a pullback is likely, and a 30-pip stop leaves little room for it.
How much do commissions cost on a candlestick pattern strategy?
On this test, 2,474 round trips at 0.02% per side add up to roughly the full position size over ten years. On an hourly chart with a 30-pip stop, costs are a large share of each trade's risk.
Reproduce it, then change it
Every number on this page comes from the Backtrex engine. Rebuild the strategy in the app, then change the market, the timeframe or a parameter and see whether the result holds.
Ready to backtest your strategies?
Create your free account and take the guided tour: a real 10-year backtest in 3 minutes, no coding, no credit card.
Create your account in 30 seconds and run your first backtest. No credit card.
Other backtests
- Strategy
- -22.1%
- Buy & hold
- +55.5%
- Strategy
- +26.9%
- Buy & hold
- +528.3%
- Strategy
- +18.1%
- Buy & hold
- +255.4%
- Strategy
- +14.4%
- Buy & hold
- +136.7%