Best trailing stop management for XAUUSD H1?

New comment
 

Hi everyone,
I am developing an EA for XAUUSD and I would like some advice about trailing stop management.

Current setup:

  • Symbol: XAUUSD
  • Mainly H1 timeframe
  • Trend/momentum strategy
  • Trades often run into profit, but many positions close too early because of trailing settings.

At the moment I am testing:

  • classic trailing stop
  • break even + trailing
  • ATR trailing
  • step trailing

My problem is finding the right balance between:

  • protecting profits
  • avoiding premature exits during gold volatility

For those trading or coding EAs on XAUUSD:

  1. What trailing logic works best for you?
  2. Fixed points or ATR-based trailing?
  3. When do you activate break even?
  4. Continuous trailing or step trailing?
  5. Any recommended ATR multipliers for H1 gold trading?

I am especially interested in real experience with volatile gold movements.

Thanks!

 
another idea: psar trailing.
 
Hi Tony,
  
  Concrete answers to your questions for XAUUSD H1 trend EAs:

  ATR multiplier: 2.5–3.5× ATR(14). Below 2× gold wicks will kill you.

  Break-even: activate at +1.5R, not 1R. Gold commonly retraces ~1R before continuing. Always leave a 10–15 pip buffer above entry — exact BE gets eaten by spread + commission.

  Step over continuous: step every 1× ATR. Continuous trailing reacts to every wick and exits prematurely on gold.

  What actually keeps you in the big runs (and is missing from your list):

  - Structure-based trailing. Trail behind the previous H1 (or H4) swing low/high once in profit. Pure ATR trailing exits on the first impulsive pullback; structure trailing rides the
  whole leg.
  - Chandelier Exit (Highest High − 3× ATR). Works very well on gold trends, better than classic ATR trailing.
  - Two-stage trail: loose ATR (3.5×) until +3R, then tighten to 1.5× ATR. Lets the trade breathe early, locks profit late.
  - Session/time filter. Tighten or close before NY close / Asian rollover if the move has run >12h — gold gives back gains overnight and spreads widen brutally at rollover.
  - News pause. Freeze the trailing during high-impact USD news (NFP, CPI, FOMC) — the spike will hit your SL and reverse 30 seconds later.

  Critical for backtesting: use real tick data + variable spread. Fixed-spread tests will overestimate trailing performance massively on XAU.

  Good luck.
 
Navdeep Singh Kaur #:
Hi Tony,
  
  Concrete answers to your questions for XAUUSD H1 trend EAs:

  ATR multiplier: 2.5–3.5× ATR(14). Below 2× gold wicks will kill you.

  Break-even: activate at +1.5R, not 1R. Gold commonly retraces ~1R before continuing. Always leave a 10–15 pip buffer above entry — exact BE gets eaten by spread + commission.

  Step over continuous: step every 1× ATR. Continuous trailing reacts to every wick and exits prematurely on gold.

  What actually keeps you in the big runs (and is missing from your list):

  - Structure-based trailing. Trail behind the previous H1 (or H4) swing low/high once in profit. Pure ATR trailing exits on the first impulsive pullback; structure trailing rides the
  whole leg.
  - Chandelier Exit (Highest High − 3× ATR). Works very well on gold trends, better than classic ATR trailing.
  - Two-stage trail: loose ATR (3.5×) until +3R, then tighten to 1.5× ATR. Lets the trade breathe early, locks profit late.
  - Session/time filter. Tighten or close before NY close / Asian rollover if the move has run >12h — gold gives back gains overnight and spreads widen brutally at rollover.
  - News pause. Freeze the trailing during high-impact USD news (NFP, CPI, FOMC) — the spike will hit your SL and reverse 30 seconds later.

  Critical for backtesting: use real tick data + variable spread. Fixed-spread tests will overestimate trailing performance massively on XAU.

  Good luck.

Hello Singh thank you so much for taking the time to write such detailed and practical advice. I really appreciate it.

Your points about the ATR multiplier (2.5–3.5×) and the break-even at +1.5R make perfect sense – gold wicks are brutal, and I've definitely seen trades get stopped out prematurely at tighter levels.

I especially liked the suggestions that go beyond a simple ATR trail:

  • Structure-based trailing behind swing lows/highs – I hadn't considered that, and it sounds much more robust for riding the full leg.

  • Chandelier Exit and the two-stage loose/tight trail – both seem very well suited for gold's impulsive moves.

  • Session/time filter and news pause – those are definitely on my list now.

Also, the backtesting note is a wake-up call. I'll switch to real tick data and variable spread to avoid overestimating performance.

I'm going to incorporate several of these ideas into my EA in the coming weeks. If you're open to it, I might ping you with a follow-up question or two once I start testing the structure-based trail.

Thanks again – your experience with XAU really shows.

Best,

 

thanks again for the detailed advice — it’s some of the most practical gold‑specific guidance I’ve seen in a long time.

A quick summary of how my trailing is currently set up (all ATR‑based, scaled on H1, 5‑digit broker):

  • Phase 1 (BE) – activates when profit reaches 0.55 × ATR. I then move the SL to entry + 0.40 × ATR (so there's a buffer, not exact BE).

  • Phase 2 – activates at 0.80 × ATR profit. SL trails behind price by 0.55 × ATR.

  • Phase 3 – activates at 2.50 × ATR profit. SL tightens to 0.35 × ATR behind price.

All steps are discrete (no continuous trailing). The initial SL is provided by two AI models (Sonnet + GPT‑4o) and typically lands somewhere between 700–1500 points (~0.7–1.5 ATR depending on volatility). If the AI doesn’t send a specific SL, the default is 0.5 ATR — and after reading your email I can already see that’s way too tight and needs to be raised.

The one idea I’m considering and would love your take on:

After TP1 is hit, I’d like to immediately move TP2 and TP3 to entry + a small offset (so BE + buffer), regardless of whether the normal Phase‑1 threshold has been reached.

The rationale is to eliminate risk on the remaining positions as soon as the first target pays for the whole trade. The offset would be 10–15 pips as you suggested, to account for spread/commission.

From your advice I understand you recommend waiting until +1.5R before activating BE, because gold often gives back ~1R before extending. If I understand correctly, my idea could be risky if TP1 is sitting at less than 1.5R — which it often is on smaller moves. So maybe the safer approach would be to only apply this “BE after TP1” rule when TP1 itself is already at least 1.5× the initial risk, otherwise let the standard ATR‑based trailing do its job.

Would you say that’s a reasonable compromise, or would you still advise sticking purely with structure‑based / Chandelier trailing and forgetting about BE‑on‑TP1 altogether?

Thanks again — looking forward to your thoughts.

Best,

Tony
New comment