forex-scalping.com

Several threads compare trading costs, but everyone uses different numbers. A common weekly format would make comparisons meaningful.

Suggested columns:

• Week, broker, account type, server location.
• Pair and session (Asia, London, NY).
• Average spread and 95th percentile spread for the session, in pips.
• Commission converted to pips per round turn.
• Average slippage on market orders, in pips, with the number of orders.
• Notes: news days, outages, rule changes.

Rules to keep it useful:

• Your own measurements only, not published averages.
• State the sample size. Five trades don't say much.
• No affiliate links, and disclose any relationship with the broker.

If a few people log a month each, we'd have a better picture than any broker comparison website. A spreadsheet template could be shared in the thread.

Would anyone be willing to log for a month? Even two or three participants on different brokers would give a first useful comparison, and gaps in the format will show up quickly once real data goes in.

I'd be up for a month, and I've actually been keeping something close to this since the spring, so I can say where the format gets awkward once real data goes in. The hardest column for me has been the 95th percentile spread. My platform only shows the live spread, so I ended up running a tiny script that samples bid and ask every ten seconds during my session and dumps them to a file. Without that, my "average spread" was really just whatever I happened to glance at.

The other thing I'd add is a column for the time of the first trade of the session. My slippage numbers looked fine overall, but when I split them by the first fifteen minutes of Frankfurt versus everything after, almost all of it sat in that early window.

Sample size is the real problem though. In a normal week I place maybe thirty market orders, which isn't much once it's split by pair. I'd suggest logging per month but reporting only once a pair has at least fifty orders behind it.