I'm looking for input with evaluating discretionary performance using average win/loss, MAE, MFE, etc.
I'm consistently making small gains daily. I have a decent win%... about 65%. I'm scalping crude (which has been terrible that last week, BTW). The concern with scalping is that the high win% is merely a function of a poor risk:reward. For example, a scalper could have 3t target and 20t stop. You could probably get away with this for a while thinking you are trading well. I imagine that an honest evaluation of MAE, MFE, ETD and average win/loss would reveal the danger of such a strategy before the inevitable big loss or two wipes out days or weeks of profits.

My average win is 4x my average loss, but average MAE is only slightly lower than average win.
I want to make sure that I am not deluding myself into thinking I have an edge or am trading well, when in reality, my results may be simply random.

How do any of you scalpers use average MAE, MFE & ETD to evaluate your trading? Would making sure that average win is larger than average MAE be a good way to assess that I am not taking too much risk?

Slipknot,

Great question. Personally I see my trading statistics as one measure of my trading "edge" similar to how a Dr. will take a patient's temperature. So is it a problem if your average MAE is nearly the size of your average winner? Maybe and maybe not. What is the reason that your system "should" have an edge (e.g., fading low vol moves, momentum breakouts, etc.)? Are your targets and stop logically placed to take advantage of your specific market? Also are you able to execute your system well enough to take advantage of that edge? Can you adapt when the market changes (like this week volatility has dropped in Crude)?

Some thoughts from an algo guy....

I would think it would be much more difficult to evaluate "strategy" effectiveness and efficiency for discretionary trading.

The biggest evaluation I think you would want to focus on is consistency and final outcome. Are you winning more than losing? Are you profitable more days than not?

Obviously discretionary "systems" are much more fluid and intangible so parameters like Maximum Adverse Effect don't really apply as much, because in a set/rigid algo method, those usually have more impact on an algo exit performance.

If you're SL ratio is 4:1 but you're P/MAE ratio is about 1:1, then that tells me you do a really good job of holding firm and letting the trade come back your way.

In short, I think you'd be better served looking at total effectiveness, bottom line and consistency....if it's working for you, and you're making money, then stick with it. I sure wish I could do it consistently (discretionary)....I don't have the discipline/stomach for it.

About the only parameter I would focus on outside of consistent returns, is MFE/Net Average Profit. That will tell you if you're leaving money on the table by staying in trades too long.

Once your consistency is there (which it sounds like it is) then the trick is simply to reduce/minimize the losers (which it sounds like you've already done) and then maximize the winners.