Drafted with AI tools. Editorial review: WickViper Team, 7 Oct 2026. Sources are listed at the end of the article.
Fails after costs
Short answer
Does a filter of “only with the current session’s direction” improve short-term gold trades?
No. The filter cut about two thirds of the trades, but the rest did the same or worse: −0.91 and −1.21 bp per trade versus −1.02 bp without the filter (2022–2024).
Data
XAUUSD, entry on an RSI(2) pullback on M5 in the daily trend
Period
2012–2024 (built to 2021, checked 2022–2024)
Sample
about 45,000 trades without the filter, about 16,000 with it
Costs
spread and commission from real trades, about 1.1 bp
Key numbers
No filter
−1.02 bp per trade
Filter: last 4 hours
−0.91 bp
Filter: day direction
−1.21 bp
Trades removed
about 64%
Why we asked
A common worry: the trend of recent weeks is down, but today’s session is pulling up — so a strategy trading the longer trend goes "against the tide" and loses. We tested it on a rule that buys short pullbacks in the direction of the daily trend.
What we tested
Two filters: (1) trade only when the price over the last 4 hours moves in the same direction as the trend, (2) trade only when the price since the day’s open moves that way. Plus variants: longs only, shorts only, both sides without the trend, against the trend only.
Result
The filters cut about two thirds of the trades, but the ones that remain are not better. On 2022–2024 the result per trade was −0.91 bp (4-hour filter) and −1.21 bp (day filter) versus −1.02 bp without one. Over the last 5 years every direction variant made −0.8 to −1.1 bp. A session against the trend is not the cause of the losses — the cause is that the rule’s own edge (about +0.3 bp) is smaller than the cost.
What it means for a trader
A filter that "makes sense" does not have to improve results. Always check that the trades left after the filter are really better, not just fewer.
Fewer trades mean lower total costs, but the result per trade does not change if the filter carries no new information.
Questions
So the session trend does not matter?
For this rule it did not improve the result. That does not mean session direction does not exist — only that this filter added no information beyond the daily trend.
What about trading both sides?
We tested that too: both sides without a trend filter give a similar result per trade but twice as many trades, so they lose faster in total.