The most durable intraday idea there is: in a trending market, pullbacks are opportunities rather than warnings. The EMA pair does nothing clever — it simply gives you an objective definition of “trending” and an objective place to wait.
Why it works
A trend persists because participants who missed the initial move keep waiting for better prices. Each pullback brings in that waiting demand, which is why trends advance in steps rather than straight lines. The moving average approximates where the trend's average cost sits, so it approximates where that waiting demand becomes active.
The strategy's real value is behavioural: it stops you buying extensions. Most intraday losses come from entering after a move has already run, where the stop must be far away and the remaining reward is small.
Market conditions required
- A visible intraday trend — higher highs and higher lows on the 5-minute chart, not a vague upward drift.
- EMAs stacked and sloping in the trend direction.
- Active session hours. This needs participation; the London–New York overlap is ideal.
- Room to the next level. If obvious resistance sits ten pips away, the reward does not justify the risk.
Setup and indicator settings
| Setting | Value | Why |
|---|---|---|
| Fast EMA | 9 | Tracks the immediate swing |
| Slow EMA | 21 | Defines the intraday trend; the band between the two is the entry zone |
| Trend filter | 200 EMA | Only take longs above it, shorts below |
| Timeframe | 5m for signals, 1m for entry | The 1-minute is for timing only, never for direction |
Entry rules
- Confirm price is above the 200 EMA and the 9 is above the 21 (reverse for shorts).
- Wait for a pullback that touches the band between the 9 and 21 EMAs.
- Require a rejection candle inside the band — a lower wick for longs, closing near its high.
- Enter on the close of that candle, or on the 1-minute break of its high.
Stop-loss rules
Below the pullback low, plus a small ATR buffer. If price closes decisively below the 21 EMA the pullback has become a reversal and the setup is void, whether or not the stop has been touched.
Take-profit rules
- T1: the prior swing high. Take roughly half. This is the highest-probability target.
- T2: trail behind 5-minute swing lows. Trends run further than expected roughly often enough to justify letting part of the position ride.
- Minimum acceptable: if T1 is not at least 1.5× your stop distance, skip the trade.
Risk management
- 0.5% risk per trade; this is a high-frequency style.
- Stop after two consecutive losses — consecutive losses usually mean the trend has ended and you have not noticed yet.
- Do not trade the same setup on three correlated pairs simultaneously.
Example winning trade
EUR/USD, 5-minute. London session, price above the 200 EMA, 9 above 21, three consecutive higher lows. Price pulls back from 1.0885 into the EMA band at 1.0868 and prints a candle with a six-pip lower wick closing at 1.0873.
Long 1.0873, stop 1.0864 below the wick, nine pips of risk. Prior swing high at 1.0885 hit for +1.3R on half; the remainder trails to 1.0898 for roughly +2.8R.
Example losing trade
GBP/USD, 5-minute. Same setup, pullback into the band, rejection candle, long taken at 1.2712 with a stop at 1.2703.
Price ticks up four pips, rolls over, closes below the 21 EMA and then below the pullback low. Stopped at 1.2703 for −1R. The trend had been running for over three hours and this was its fifth pullback — late-stage trends fail on pullbacks, which is exactly how they end.
When NOT to trade it
- Ranging or chopping markets where the EMAs are flat and intertwined.
- The fourth or fifth pullback of an extended trend — the risk-reward degrades with each one.
- During lunchtime lulls when volume disappears.
- Into major news, where the trend structure is about to be overwritten.
Common mistakes
- 1Trading it in a range
Without a trend there is nothing to continue, and every touch of the band is noise.
- 2Skipping the rejection candle
Buying merely because price reached the EMA removes the only confirmation the setup has.
- 3Chasing the entry
If price has already left the band, the trade is gone. Another will come.
- 4Trading every pullback in a tiring trend
Fewer, better entries beat more entries. Track pullback number in your journal and you will see the decay.
Backtesting considerations
- Record which pullback number each trade was; expectancy typically falls after the third.
- Test with the 200 EMA filter on and off — the filter usually accounts for most of the edge.
- Costs matter enormously at this frequency; a strategy that nets 6 pips before costs and pays 1.5 in spread is a very different proposition.