Page 3 of 3
Re: Correlated Pairs = Correlated Risk
Posted: Mon Sep 21, 2026 12:59 pm
by PropScalpDesk
Fairman wrote:Running simultaneous scalps on EUR/USD and GBP/USD at the same time might feel like diversification. It isn't.
USD × 2 wearing two tickets. If each is “1%,” a strong dollar impulse can behave like one larger bet with two spreads. I name the shared idea first; then I cap aggregate risk across correlated expressions. One primary usually keeps the full seat; the second is reduced or watch-only.
Correlation shifts, so I recheck rather than trusting last year’s textbook pairs. EURUSD/USDCHF offsets are not a free hedge if the session is pure risk-on chaos. On overlap I am especially strict — that is when stacking feels smart and fails together.
Funded flavour: trailing DD turns correlated heaters into limit events. Better one clean R than two “diversified” scratches that flip in the same minute.
Weekly I glance at rolling correlation, not a textbook memory. When risk-on chaos hits, even supposed offsets can travel together long enough to finish a soft daily stop.
Do you use a hard aggregate % cap, or a simple “one USD idea at a time” rule?
Re: Correlated Pairs = Correlated Risk
Posted: Sat Oct 03, 2026 10:34 am
by Fairman
PropScalpDesk wrote: Do you use a hard aggregate % cap, or a simple “one USD idea at a time” rule?
One USD idea at a time. It's easier to follow under pressure than a percentage cap. If I'm long EURUSD I'm not also long GBPUSD, I just pick the cleaner chart.
Re: Correlated Pairs = Correlated Risk
Posted: Sat Oct 03, 2026 5:56 pm
by LondonNewsTrader
PTScalper wrote: Here is the professional-grade adaptation for cTrader (Automate / cBot) written in C#.
There are some issues in CalculatePearson that will change the numbers. The index is computed from bars1.Count and then used on bars2 as well. Two symbols rarely have the same amount of history loaded, so ClosePrices[index] on the second series usually refers to a different hour. Each bar needs to be matched by open time.
It also correlates price levels rather than returns. Two trending pairs will show a high correlation whether or not they actually move together from bar to bar, so I'd use the change in close.
The volume filter compares VolumeInUnits with MinLotSize × Symbol.VolumeInUnitsMin. VolumeInUnitsMin is the minimum order size, typically 1,000 units, so 0.5 × 1,000 is 500 units and every position passes. Symbol.QuantityToVolumeInUnits(MinLotSize) gives the intended threshold, ideally using each position's own symbol.
Finally, the direction of each position matters. Long EURUSD and long USDCHF have strongly negative price correlation, but as positions they partly offset each other. Multiplying each pair's correlation by the product of the position signs gives a figure that reflects the actual risk.
Re: Correlated Pairs = Correlated Risk
Posted: Mon Oct 05, 2026 12:38 pm
by PropScalpDesk
One USD idea at a time is easier under pressure than a percentage spreadsheet. Long EURUSD means I am not also long GBPUSD — I pick the cleaner chart.
Two tickets that share a dollar impulse are one bet with two spreads.
Re: Correlated Pairs = Correlated Risk
Posted: Sat Oct 31, 2026 10:36 am
by FXS Prop Desk
The doubled-trade point is clear and important. One way to handle it in practice is to size by currency exposure rather than by pair. If EURUSD and GBPUSD are both long, the account is short USD twice, so the combined risk should stay at the 1% planned for one idea. Splitting it as half a percent on each pair keeps the total under control. The correlation is not constant, though, so on days when GBP has its own news the two pairs can separate.