This is an interesting thread... to simplify matters, lets take the example of a 1-lot ES trader... the S&P currently has a value of 933, so at $50 a point an ES contract has a face value of $46,650.
Now let's assume the 1-lot ES trader starts with $5000 in his account, which comprises enough for intraday margin and a buffer for drawdowns.
Now, lets assume that our 1-lot ES trader makes 20 points net in his first month (let's further assume that our guy is not a newbie, but has experience of trading stocks, but is dabbling in ES to get a feel for it alongside his primary stock trading)... given these assumptions, I assert that a point a day net of trading costs is highly achievable in ES, being a modest and realistic assumption... so, at the end of the month, our 1-lot ES trader now has $6000 in his account...
How do we measure his performance?
Lets dig deeper... lets assume that the ES trader had at all times used a 0.75 point stop loss, since he is a scalper... $37.50 is the value of his stop, which is 0.75% of his initial capital...
We now have several ways to measure his performance:
1) $1000 / $5000 = 20% return on initial deposit
2) $1000 / $46,650 = 2% return on initial face value of the contract
3) the actual dollar profit of $1000
4) the net points of profit, which is 20
5) 27x the average stop loss of 0.75 points
(Please note that the above numbers are merely approximations, since further assumptions would have to be made about position sizing, trade management etc... but for simplicity, we will regard these numbers as broadly acceptable...)
I am inclined to believe that 5) is the best measure of trading acumen, since trading performance should be a reflection of how you are compensated for bearing risk, as opposed to being contingent on a somewhat arbitrary combination of margin choice and start-up capital... in my view, we need to move away from the convention of % returns and/or absolute $ values, if we are to arrive at a consistent yardstick for measuring trading performance...