Gold exits: why my ATR-based TP almost never got hit, and what I changed

This is where I lost the most time on gold (XAUUSD), so I want to share what happened and hear how you handle it.

What I tried first

• SL = ATR x factor, TP = a fixed multiple of the SL (the classic 1:2 / 1:3).
• On gold the ATR is large, so the SL was wide, and the TP ended up even further away.
• Result: price very often moved in my favour, then came back. The TP was almost never reached, and many trades that were in profit closed at a loss or at breakeven.

What I changed

• I stopped using ATR as the unit for SL/TP. I still use volatility as a filter (trade or not), but the SL and TP are now a fixed % of price.
• I moved toward a higher win rate with smaller, realistic targets. I watch two numbers closely: the size of the biggest single loss and the longest losing streak. A high win-rate system can be killed by one big loss.
• I compare every exit idea on the same trades with only the exit changed, not with a completely new backtest, so I can see what the exit itself does.

Questions for you

1. On XAUUSD, do you set SL/TP from ATR, from % of price, from structure (swing high/low), or something else?
2. Has trailing ever really helped you on gold, or did it mostly give back profit?
3. How do you judge whether one exit is better than another: profit factor, MFE/MAE, win rate, max losing streak?
4. Do you use different exits for the Asian session and for London/New York?

Please share in the comments what actually worked or failed for you, in testing or live. I will add a summary of the answers to this post.