forex-scalping.com

I had raw accounts at both for most of the summer and traded gold on each in roughly equal size, so I finally have a cost log worth posting. Please read this as one trader's sample, not a statement of either firm's pricing. Spreads move and commission schedules change, so check current terms yourself.

The sample: 212 XAUUSD round trips on one account, 197 on the other, all between 07:00 and 16:30 London, June to early September, sizes of 0.3 to 1.0 lots.

What I logged for each ticket was the spread at the moment I clicked, the fill price versus the price I clicked, and the commission charged. I then converted everything to dollars per lot round trip so the two accounts could be compared directly.

Headline from my own numbers: the difference in all-in cost per lot was small, under a dollar on the median ticket. The bigger differences were by time of day, not by broker. Both accounts were noticeably more expensive in the 13:25 to 13:40 window on US data days and cheaper from 08:30 to 11:00.

My conclusion for my own trading is that choosing my hours carefully saves me more than choosing between those two. I've kept both open for now, mainly as a backup in case one platform has an outage.

Logging the spread at the moment of click rather than an average is what makes this useful. Most comparisons I see use advertised averages, which hide the window effect you found completely.

The 13:25 to 13:40 result fits my own log on gold. On US data days the all in cost in that window was often two to three times the morning cost on any account I've used, so the broker choice mattered less than whether I was trading then.

One field I'd add if you keep logging: side of the trade. On gold I found slippage on buys and sells wasn't symmetrical during fast moves, and the average hid that.

Also worth separating the fills during the first five minutes of London from 08:30 onward. In my numbers those first minutes were the second most expensive window after US data.

Did your fill versus click numbers show more negative slippage than positive, or roughly even?

A different angle on a cost log like this: the size effect. You traded 0.3 to 1.0 lots, and I wonder whether the cost per lot held flat across that range or started to rise near the top.

On gold I saw a step at around 1 lot where fills during fast moves got noticeably worse, about 5 to 10 cents more per ounce on average. Below that the cost per lot was basically constant.

If that pattern is common, it means a comparison at small size might not hold for someone trading 3 lots, and the reverse. The backup account point is underrated too. Having a second platform logged in has saved me more than any cost difference.

The other thing I'd measure is rejections and requotes, if either account ever gave them. A cheap fill that sometimes doesn't happen at all has a cost that won't show in the per lot number.

Did either account ever refuse an order or requote during the US data windows?

LondonScalper wrote: the fill price versus the price I clicked

How were stop-loss exits handled in the log? The click price works as a reference for market entries and manual closes, but a stop isn't clicked, so its slippage would need the stop level as the reference instead. On gold that's usually where the real cost difference between brokers shows up, since stops tend to trigger in fast moves and some venues fill them further away than others. If the 212 and 197 round trips include stopped trades measured against the click price of the entry, the comparison might mostly reflect the quieter part of the costs. Did you split stop exits out, and did the under a dollar gap still hold for those?