Quick Answer
The most effective XAUUSD EA indicator stack is EMA 200 (trend filter) + ATR 14 (volatility-based stop sizing) + one entry indicator such as RSI or ADX. This three-indicator foundation covers trend direction, risk management, and entry timing without the over-fitting risk that comes with stacking 6+ indicators.
EA Indicator Stack Builder
Toggle indicators on and off to see how your stack scores. A well-built stack has clear roles for each component.
Stack Health — 0/6 indicators active
Warning: no trend filter
Why EA Indicators Work Differently to Manual Use
When a trader manually uses indicators, they apply judgment alongside the signal. They might see an RSI reading of 72 and decide that while it suggests overbought conditions, the strong trend means the trade is still valid. They consider context, news, and recent price action.
An EA applies indicator logic literally. If RSI is above 70, the EA will reject the trade — regardless of whether a strong trend makes it the right decision. This rigidity is both a strength (consistent execution) and a weakness (inability to adapt to context). The implication is that indicator thresholds inside an EA need to be set more thoughtfully than in manual trading, because there is no human override.
This is also why fewer, simpler indicators usually outperform complex stacks in automated systems. Each additional indicator creates another condition that must be met for a trade to fire. More conditions means fewer trades, and in live markets the conditions are never exactly what they were in the backtest.
What Each Indicator Contributes to an EA
EMA 200
The EMA 200 on H4 defines whether the EA should look for long or short trades only. Trading in the direction of the EMA 200 is the single highest-leverage improvement most gold EAs can make. Without it, the EA takes trades in both directions regardless of macro context, and roughly half those counter-trend trades will be fighting a wall.
ATR 14
ATR (Average True Range) measures how much XAUUSD typically moves in a given period. Using ATR to set the stop loss — rather than a fixed pip value — means the stop scales with current volatility. A 1.5× ATR stop is realistic in low-volatility conditions and not unreasonably tight in high-volatility conditions, giving the trade room to breathe.
ADX 14
ADX measures how strong the current trend is, regardless of direction. Many EAs add a filter: only take trades when ADX is above 20 or 25. This prevents the EA from entering during sideways, choppy markets where breakout signals produce many false starts. ADX above 30 generally indicates a strong trend where breakout strategies perform best.
RSI 14
RSI in a gold EA is best used for entry confirmation rather than as a primary signal. A breakout trade with RSI momentum in the same direction is a stronger entry than a breakout against the RSI reading. Common settings: enter longs when RSI crosses above 50, enter shorts when RSI drops below 50 — keeping the threshold neutral rather than at overbought/oversold extremes.
Bollinger Bands 20/2
Bollinger Bands provide a dynamic view of price range. A close outside the upper band confirms a breakout has statistical significance — price has moved more than two standard deviations from the recent mean. Used in this way, Bollinger Bands add confirmation to a range breakout entry rather than being a standalone trading signal.
MACD 12/26/9
MACD captures momentum changes through the relationship between two EMAs. In a gold EA, the MACD histogram crossing zero can confirm the start of a new momentum phase. It is most useful on H1 for medium-frequency systems; on M5 or lower, MACD generates too much noise to be a reliable filter.
How Over-Optimisation Kills EA Profitability
Over-optimisation is one of the most insidious problems in EA development. The process of backtesting and tweaking parameters can produce an EA that performs spectacularly on historical data but fails in live trading. This happens because the optimised parameters are precisely fitted to the patterns in the historical data — patterns that will not repeat exactly in the future.
Adding more indicators compounds this problem. Each indicator adds parameters that can be optimised. RSI period, ADX threshold, EMA length, ATR multiplier, MACD settings — an EA with all six indicators might have 15–20 tunable parameters. An optimizer can find a combination of these that produces an equity curve that looks perfect on 2022–2025 data, then completely fails in 2026.
Red flags for over-optimised EAs
!Backtest win rate above 85% (markets are not that predictable)
!Maximum drawdown under 3% over multiple years (too clean)
!Performance degrades sharply when tested on a different date range
!Requires very specific indicator settings (e.g., RSI period of 17 or ATR multiplier of 1.73)
!Large performance gap between backtested and live results in first month
Goldie Razor V2.8.4 uses EMA 200 and ATR-based stop sizing as its core indicator logic — a lean stack that avoids over-fitting while covering the two most important aspects of gold EA trading: trend direction and volatility-adjusted risk control.
Related Questions
Frequently Asked Questions
Goldie Razor V2.8.4
M15 breakout + H4 EMA filter — built for XAUUSD on MT5