September scalp review: what I change for next month

[LondonScalper]

September scalp review: what I change for next month

Monthly reviews are where I stop lying to myself about “a few bad days.” September’s book usually shows the same sins in new clothes: late-NY overtime, gold size drift, and news curiosity dressed up as a plan. Without a fixed template I drift into storytelling.

My September template stays fixed year to year:

• Cost per R by pair — not just win rate.
• Which session produced expectancy after costs.
• Which rule I broke most often (process score, not P&L).
• One change only for next month. Not five “improvements.”

Last useful change from a September pass: earlier hard shut-off on Fridays. Give-back trades vanished from the log within two weeks. The discipline is choosing a single lever and measuring it, not redesigning the whole playbook because the month felt annoying.

If you do a monthly scalp review, what is the one change you are carrying into next month?

I keep the September file next to the prior two months so I can see whether the same rule-break keeps mutating. If the same sin appears three months running, the next change is environmental — earlier shut-off, smaller gold envelope — not another entry tweak.

[PropScalpDesk]

One change for October from me, and it's a calendar one. Europe leaves summer time on 25 October, the US only on 1 November. For that week, US releases arrive an hour earlier on European clocks, so the 13:30 London prints move to 12:30. Every year somebody gets caught by a big US release during what they thought was lunch.

So my October change is just re-checking every blackout time for that week by hand instead of trusting last month's sheet.

On the method, one change per month is right, but it only works if you judge it on the same metric you picked it for. Your Friday shut-off is a good example: count of give-back trades before and after, not the monthly P&L, which has a dozen other things in it.