mean reversion on gold, does it actually work or is gold a trender

[Dean]

reading mixed opinions on whether gold (XAUUSD) is a trender or a mean-reverter. classical wisdom: gold trends hard in macro regimes (gold rallies during inflation, sells off during disinflation). but intraday and weekly, gold shows very clear mean reversion to its moving averages.

i've been testing a mean reversion strategy on gold for 3 months: short at 2 ATR above 20EMA, long at 2 ATR below. modest win rate ~58%, but 1:1.2 R on average. profitable but not exciting.

curious what others have tested or what their experience is with gold's character.

[bluedreams]

gold is both, in different timeframes simultaneously. weekly chart: trender, driven by macro. 4H/1H: mean reverter, prices oscillate around moving averages. daily: hybrid, depends on regime. the trick is matching strategy timeframe to character expressed at that timeframe.

[endlessroad]

exactly. i run a mean reversion strategy on 1H gold with positive expectancy. but i ALSO have a weekly trend follower on gold. they're not contradictory - they capture different aspects of price behavior at different timeframes.

[Gavin]

isnt it confusing to run both at the same time? what if one says long and the other says short?

[silentcore]

they often will conflict and thats fine. you treat them as separate accounts effectively. weekly trender holds a long-term long, intraday mean reverter takes a short for 4 hours, both work because they harvest different inefficiencies. just make sure your aggregate position sizing accounts for net exposure.

[Maverick]

tested mean reversion on gold across 3 years of 1H data. results vary wildly by regime. in trending years (2019-2020) mean reversion drawdowns were brutal. in choppy years (2022-2023) it printed money. on average across all years it was profitable but volatile. needs a regime filter to be reliable.

[justmatt]

gold is whatever you need it to be to justify your strategy that month. confirmation bias is real.

[bluedreams]

i strongly disagree that gold is a mean reverter at any meaningful timeframe in 2026. since central banks started accumulating heavily, gold has been in a structural macro uptrend with mean reversion only as noise around the trend. the rules of pre-2020 dont apply.

[samuel.j]

gold is my favorite instrument now i dont care if its trending or reverting on any given day, the volatility means there are setups every session. just match strategy to timeframe and stay disciplined

[Cody]

timeframe dependent. respect both.