- Collect spread every 10 seconds for a few weeks on each account, using the same VPS.
- Compute the average and the 95th percentile for the hours you actually trade.
- Add commission converted to pips per round turn.
- Place a set of micro market orders at your hours and record slippage.
- Check stop level, freeze level and server location.
Two accounts can look identical on average spread and still differ in the 95th percentile, which is when your stops and news entries fill. That number deserves its own weight.
Run the comparison again after any broker announcement about pricing changes.
What's the biggest difference you've found between two raw accounts?