Three Fast EMA Crossover Strategy: Entry, Exit, and Filter Rules
The three fast EMA crossover uses three exponential moving averages for rapid trend entries. Learn the 5/10/20 EMA setup, the dual-crossover confirmation rule, and how to filter false signals in trending conditions.

The three fast EMA crossover is an evolution of the standard two-EMA crossover system. By adding a third, intermediate EMA, the strategy requires two crossovers to confirm a directional signal — the fast crossing the intermediate, and the intermediate crossing the slow — significantly reducing false entries in choppy markets.
It’s more selective than a single crossover system, produces fewer signals, and filters out many of the whipsaws that make basic crossover strategies frustrating in ranging conditions.
The Three EMAs and Their Roles
| EMA | Period | Role |
|---|---|---|
| Fast EMA | 5 | Fastest response to price; first crossover trigger |
| Medium EMA | 10 | Intermediate confirmation; second crossover trigger |
| Slow EMA | 20 | Trend direction baseline |
The alignment that confirms a trend:
- Bullish alignment: 5 EMA > 10 EMA > 20 EMA (all three stacked in order, all sloping up)
- Bearish alignment: 5 EMA < 10 EMA < 20 EMA (all three stacked below, all sloping down)
Full bullish alignment is required before entering long. Full bearish alignment is required before entering short.
Entry Rules
Long Entry (All Conditions Required)
| Condition | Requirement |
|---|---|
| 1. Daily trend | Price above 50 SMA on daily chart |
| 2. Fast cross | 5 EMA crosses above 10 EMA on 4H chart |
| 3. Full alignment | 5 EMA > 10 EMA > 20 EMA on 4H chart |
| 4. Candle close | Current 4H candle closes with all three EMAs stacked bullish |
| 5. Entry | Market order at open of the next 4H candle |
Short Entry (All Conditions Required)
| Condition | Requirement |
|---|---|
| 1. Daily trend | Price below 50 SMA on daily chart |
| 2. Fast cross | 5 EMA crosses below 10 EMA on 4H chart |
| 3. Full alignment | 5 EMA < 10 EMA < 20 EMA on 4H chart |
| 4. Candle close | Current 4H candle closes with all three EMAs stacked bearish |
| 5. Entry | Market order at open of the next 4H candle |
The “full alignment” requirement is the key filter. If the 5 EMA crosses the 10 EMA but the 10 EMA is still above the 20 EMA (for a short), full bearish alignment doesn’t exist yet. Wait.
Exit Rules
Stop-Loss
Place the stop-loss below the 20 EMA at time of entry (for long), or above it (for short), plus a 10–20 pip buffer:
Long stop = 20 EMA price at entry - 15 pips
Short stop = 20 EMA price at entry + 15 pips
Alternatively, use the most recent swing low/high as described in the chart stop-loss lesson. Whichever produces a wider stop gets priority (the one that gives the trade more room).
Take-Profit Options
Option A — Fixed RR: Target 2× the stop-loss distance.
Option B — EMA Trailing: Exit when the 5 EMA crosses back below the 10 EMA (long) or above it (short). This lets winners run in strong trends.
Option C — Hybrid: Close 50% of position at 2× risk; trail the remaining 50% using Option B until the 5/10 EMA reversal cross.
In strongly trending markets, Option B (or the hybrid) typically captures significantly more profit. In choppy, slow-moving markets where trends reverse frequently, Option A provides more reliable results.
Optional Additional Filters
RSI filter: Only take long entries when RSI (14) is above 50; only take short entries when below 50. This ensures entries align with the momentum direction.
ADX filter: Only trade when ADX (14) is above 20. Below 20, the three EMAs will stack and unstack repeatedly in a ranging market, producing many false signals.
Volume filter (if available on your pair): A crossover on above-average volume is more reliable. For forex pairs without direct volume, use tick volume as a proxy.
Comparison to Two-EMA Systems
| Feature | 2 EMA System | 3 Fast EMA System |
|---|---|---|
| Signal frequency | Higher | Lower (more selective) |
| False signal rate | Higher in ranging markets | Reduced by full alignment requirement |
| Entry timing | Earlier | Slightly later (requires full stack) |
| Best in | Strongly trending markets | Both trending and moderately ranging |
| Trailing exit | Simple single cross | More nuanced (5/10 cross exit) |
Risk Management
| Parameter | Recommendation |
|---|---|
| Risk per trade | 1% of account |
| Stop-loss | 20 EMA + 15-pip buffer or swing structure |
| Minimum RR | 1:2 |
| Max simultaneous trades | 2 |
| Avoid | Trading within 30 minutes of major news events |
Use the position size calculator for exact lot sizing.
When This Strategy Underperforms
This system is inherently trend-following. It underperforms when:
- The market is in a defined range (price oscillates without directional momentum)
- ADX is below 20 (no meaningful trend)
- After a major news event creates a temporary spike that quickly reverses (the EMAs stack briefly then immediately unwind)
Adding the ADX filter eliminates most underperformance in ranging conditions. The daily trend filter (50 SMA) eliminates many counter-trend signals on the 4H chart.
Key Takeaways
- Three fast EMAs (5/10/20) require full bullish (5>10>20) or full bearish (5<10<20) stacking for an entry signal — not just the first fast-over-medium cross
- Daily chart 50 SMA direction must align with the 4H signal direction
- Stop-loss below the 20 EMA (long) or above it (short) with a pip buffer
- Exit options: fixed 1:2 RR, 5/10 EMA reversal cross (trailing), or hybrid (50% fixed, 50% trailing)
- Add ADX >20 filter to avoid false signals in ranging markets
- Generates fewer signals than a two-EMA system but with meaningfully fewer false entries
Related strategies: EMA Crossover and ADX | Moving Averages Crossover
Related lessons: Moving Averages
Tools: Position Size Calculator
Risk Disclosure & Testing
Past performance of any trading strategy is not indicative of future results. Always backtest and forward test this strategy on a free demo account before committing real financial capital.
Practice This Strategy on a Live Demo
Choose an ECN broker with tight spreads and fast execution for strategy backtesting.