How to better filter false entries in trend-following strategies on XAUUSD during volatile periods?

[Wei Xin Sun]

Hello everyone in the Trading Systems section!

I'm trying to improve a simple trend-following approach specifically for XAUUSD (Gold), and one big issue I'm facing is too many false signals/entries during high volatility (news events, sudden spikes, choppy ranges).

What I've tried so far:

Using a higher timeframe filter (like D1 to confirm overall trend direction).

Adding some basic volatility check to skip entries when market is too wild.

Strict risk rules (fixed SL/TP, daily drawdown cap).

But in backtests (using Every tick based on real ticks where possible), false breakouts still cause unnecessary losses or whipsaws.

Questions for the community:

What additional filters or conditions do you find most effective to reduce false entries on gold in volatile times? (e.g., ATR-based, time-of-day restrictions, news avoidance logic, or something else?)

How do you balance "not missing good trends" vs "avoiding bad trades"? Any favorite indicators or code ideas?

In Strategy Tester, do you always trust "Every tick based on real ticks" results for gold, or do you cross-check with different modes/brokers?

I'm open to any suggestions, code snippets, or experiences you've had with similar setups. No promotion here — just trying to learn and make better decisions in automated trading.

Thanks a lot for any input! Looking forward to your thoughts.

[[Deleted]]

You can reduce false entries on gold by combining time‑of‑day restrictions with volatility filters like ATR or minimum candle structure confirmation. Balancing missed trends and bad trades works best when you keep one core entry logic but require multiple lightweight confirmations so the system stays flexible rather than over‑filtered.

[Wei Xin Sun]

Petra5 #:
You can reduce false entries on gold by combining time‑of‑day restrictions with volatility filters like ATR or minimum candle structure confirmation. Balancing missed trends and bad trades works best when you keep one core entry logic but require multiple lightweight confirmations so the system stays flexible rather than over‑filtered.

Hi [Petra5 #]，

Thank you so much for the detailed reply! This is exactly the kind of practical advice I was hoping for.

I like the idea of combining time-of-day restrictions (e.g., avoiding high-impact news windows or low-liquidity Asian sessions) with ATR-based volatility filters — it makes sense for gold's "spiky" behavior. The "multiple lightweight confirmations" approach also resonates; I've been over-filtering in some tests, which reduced trades but sometimes missed strong trends.

A few follow-up questions if you don't mind:

1. For ATR filter, what period and multiplier do you typically use on XAUUSD (e.g., ATR(14) > 1.5x average for skipping entries)?

2. What kind of "minimum candle structure" confirmation works best for you? (e.g., engulfing patterns, pin bars, or just body size > certain % of range?)

3. Have you tested this on different brokers/spreads? Gold spreads can vary a lot and affect false signal filtering.

Thanks again — your input is helping a lot! Looking forward to more ideas from the community.

[Shanmugi]

Wei Xin Sun #:
1. For ATR filter, what period and multiplier do you typically use on XAUUSD (e.g., ATR(14) > 1.5x average for skipping entries)?

ATR(M1, 2) >= ATR (M1, 1440) x 2 (Do not take trade) 

it comparing RealTime ATR Value with Last 24 hours average. 

[Osmar Sandoval Espinosa]

Perhaps this articles also helps (it's based on ICT concepts):

Optimizing Liquidity Raids: Mastering the Difference Between Liquidity Raids and Market Structure Shifts - MQL5 Article