EMA 21 + RSI momentum
NAS100, XAUUSD · 5m chart · strategy s19_ema21_rsi · 10 variations
Short-term momentum trades on 5-minute gold and NAS100 charts: it enters when price and momentum turn together, then rides the move with a wide trailing stop.
What it is about
Two common indicators do the work. The EMA 21 is an average of the last 21 candles that gives more weight to recent ones; the RSI 14 (Relative Strength Index) is a 0-100 gauge of momentum, with 50 as the middle. When, on the same candle, price closes across the EMA and the RSI crosses 50 in the same direction, price and momentum agree that the short-term direction has turned.
That alone is not enough: the next candle must break past the signal candle's high (for buys) or low (for sells), proving the move has follow-through. The trade then has no fixed target; once it is 3R up (R = the amount risked), the stop follows 3R behind the best price, so the occasional big intraday move is kept.
On 5-minute charts the broker's spread is a large part of each trade, and the rules exactly as shown in the video lost money on every market. Two changes make it work in the backtest: trading only in active hours, and skipping any trade whose stop is smaller than 10 times the spread ($3 on gold, 15 points on NAS100).
How it trades
Intraday only. Gold: the New York morning, 08:00-12:00 (I, J), or the Asian evening, 18:00-03:00 (B). NAS100: 09:00-12:00 (C), 08:00-11:00 (D) or 03:00-06:00 New York, the London morning (E). Between about 9 and 42 trades a month depending on the variation.
- Roughly 1 trade in 3 wins (27-38% depending on the variation). Most trades lose about 1R; the winners that reach the trailing stop are often worth several R.
- Main gold version (I, New York morning), backtest 2018-today at zero cost: 1,514 trades (about 14 a month), 35% wins, +333.9R, largest drawdown about 29.5R; all 9 years positive. Longest losing streak 17 trades (J, with the half-off rule: 11).
- Drawdowns are sizeable for the number of trades: about 24.4R to 47.7R across the variations.
- On gold, most trades and most of the profit come from 2024 onwards, when gold's bigger candles let more stops pass the $3 minimum.
- Every trade is closed by the end of the day; nothing is held overnight.
- The time windows, minimum stops and filter thresholds were chosen while looking at the whole history, so there is no clean out-of-sample test. Paper trading is the real test.
- Gold's $3 minimum stop makes the history before 2024 thin: the Asian-session version had only 268 trades in 2018-23 against 1,118 since, so it mostly reflects the volatile 2025-26 market.
- Very sensitive to costs: without the minimum stop the spread eats the same rules on every market. Check your broker's spread in the calculator.
- Market behaviour has shifted: big signal candles paid well before 2024 and have failed since, which is why I skips signal candles over 2.5 ATRs; what works now may not last.
- The NAS100 versions C and D are weak in 2026 so far (C flat at 0.0R, D +19.1R).
Where it comes from
From a trading video Victor shared on 2026-10-14, tested on gold, NAS100 and US500 and then built into the engine. Because settings were picked on the full 2018-2026 history, the 2018-23 vs 2024-26 split is shown for information only; every variation is being paper-traded before it is trusted. In October 2026 the gold New York versions A, F, G and H were retired: they took 57-99% of the same trades as I (which already combines F's gap rule and H's candle cap).
Main variation: s19-D ★ Main
The variations
Each variation is the same strategy with a different approach, trading its own paper account. Click one to see exactly what it changes.
NAS100 from the pre-market into the first 90 minutes of the cash session, with the 3R trailing exit and no minimum stop. Since 2026-10-21 it skips setups whose stop is wider than 35 points: our loss study found the widest stops lost in every period.
s19-AXAUUSD ·
s19-BGold, Asian eveningSame rules, but entries between 18:00 and 03:00 New York, the Asian session. Gold usually moves less then, so few setups clear the $3 minimum stop except in volatile periods.
s19-CNAS100 09:00-12:00, RSI exit + ranging filterNAS100 in the first part of the cash session, with a 15-point minimum stop. It exits when the RSI reaches 70 / 30 instead of trailing, and skips choppy markets where price crossed the EMA more than 3 times in the last 20 candles.
s19-ENAS100, London morning 03:00-06:00NAS100 between 03:00 and 06:00 New York, when London is open and New York is asleep, with a stop between 15 and 35 points (the 35-point cap was added on 2026-10-21 after a loss study) and the 3R trailing exit. A different time of day spreads the risk away from the New York versions.
s19-FXAUUSD ·
s19-GXAUUSD ·
s19-HXAUUSD ·
s19-IGold morning, gap rule + 2.5 ATR capGold, entries 08:00-12:00 New York with the $3 minimum stop and the 3R trailing exit, plus two filters: Victor's fair value gap rule (no sells inside an active bullish 5-minute gap zone, no buys inside a bearish one) and a cap that skips signal candles taller than 2.5 ATRs (big crossing candles paid before 2024 and have failed since). It replaced the older gold morning versions A, F, G and H, which took mostly the same trades.
s19-JGold morning, as I + half off at 1.5RExactly I's trades, but half the position is closed at +1.5R and the rest rides the 3R trailing stop. Victor's request to shorten losing streaks: trades that reach +1.5R and then fall back are no longer full losses.
Variations side by side
Live result of all 10 together: −16.1R (zero cost).
| Live paper account (R, zero cost) | Backtest 2018 – today (R, zero cost) | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| s19-D★ MainEMA 21 + RSI momentum | NAS100 | 5m | −2.0R | 2 | 0% | – | +713.2R | +0.26R | 8/8 | 29.4 | 28% | −47.7R |
| s19-As19_ema21_rsi | XAUUSD | −4.0R | 4 | 0% | – | – | – | – | – | – | – | |
| s19-BEMA 21 + RSI momentum | XAUUSD | 5m | −8.0R | 29 | 21% | 1 | +277.8R | +0.19R | 9/9 | 13.6 | 28% | −44.1R |
| s19-CEMA 21 + RSI momentum | NAS100 | 5m | +4.8R | 13 | 46% | – | +231.5R | +0.16R | 7/8 | 15.1 | 38% | −29.7R |
| s19-EEMA 21 + RSI momentum | NAS100 | 5m | +1.8R | 11 | 27% | – | +350.8R | +0.26R | 8/8 | 14.5 | 28% | −31.0R |
| s19-Fs19_ema21_rsi | XAUUSD | −4.0R | 4 | 0% | – | – | – | – | – | – | – | |
| s19-Gs19_ema21_rsi | XAUUSD | +0.6R | 5 | 40% | – | – | – | – | – | – | – | |
| s19-Hs19_ema21_rsi | XAUUSD | −3.0R | 3 | 0% | – | – | – | – | – | – | – | |
| s19-IEMA 21 + RSI momentum | XAUUSD | 5m | −4.2R | 21 | 19% | – | +320.9R | +0.21R | 9/9 | 14.6 | 35% | −29.5R |
| s19-JEMA 21 + RSI momentum | XAUUSD | 5m | +1.9R | 14 | 43% | – | +234.0R | +0.15R | 9/9 | 14.6 | 44% | −24.4R |
Main = the variation with the most backtest R per R of max drawdown (2018 – today, zero cost). It is picked on that history, so the live accounts are the real test.