forex-scalping.com

I keep broker decisions in a spreadsheet because memory is loyal to the last good fill.

Columns that earn their keep: median spread at 08:00, 13:00 and 15:00 London on EURUSD and GBPUSD, reject count on stops, days taken for a small withdrawal, and a notes field for platform quirks. I do not score marketing claims. I score what my deals tab shows over two weeks. Colour only marks fail thresholds I set in advance, for example more than two unexplained rejects in a London week.

When two brokers tie, withdrawal behaviour and weekend stability break the tie. The sheet is dull on purpose. Dull is how I scale size without romance.

Template layout if anyone wants to copy it: one tab per broker, one row per trading day, and a summary tab that only pulls medians. I avoid averages because one ugly minute around CPI can wreck an average and tell you nothing about the 08:15 entry you actually take every day. Twice a year I delete brokers I no longer use so the sheet stays readable.

What columns would you add? I have been thinking about a swap column for the rare overnight hold.

Median columns beat averages on my sheet for the reason you named. One ugly CPI minute wrecks a mean and says nothing about the 08:15 EURUSD entry I actually take most days.

Reject count on stops with a fail threshold set in advance is the column that has saved me from romantic loyalty to a path. More than two unexplained rejects in a London week paints red and freezes size increases.

Withdrawal days and weekend stability break ties when medians look similar. Platform quirks live in notes, never as a scored marketing claim.

A swap column for the rare overnight hold is on my add list too. I would also add a server time offset check once a month so daylight saving cannot silently skew the hour buckets.

Which column would you add first beyond spreads and rejects, and do you keep one tab per broker the same way? Dull tabs still beat memory for me.

LondonScalper, the column I would add first is slippage split by direction. On stop orders I log fill price minus requested price, with positive and negative kept apart, because a broker can show a tidy median and still be lopsided.

One feed I tested had a median slippage of zero on EURUSD stops, which looked perfect. The split showed 31 fills against me and only 9 in my favour over the two weeks, and that asymmetry is exactly what a median hides.

I would also put a 90th percentile next to each median spread. The median tells me the normal cost of a routine entry, while the 90th percentile tells me what a bad minute costs inside the hours I actually trade.

A minimum fill count stops the sheet from scoring noise. Under about 40 fills in the window, the cell stays grey instead of green or red.

Limit orders get their own rows. Mixing them with stop fills drags the slippage numbers toward zero and flatters everyone.

Where do your spread readings come from, platform snapshots at fixed times or the spread recorded on each fill? Mine come from fills, which skews toward the minutes I chose to trade.