I'm confused by what you said. You said averaging in isn't for you, but then say markets are mean reverting. Then you say that when you lose you take the max loss, but then when you get the right spot, you have a small position (as if to complain about small wins). Did you mean to say that you want to be averaging in?? All your rationale points to this answer.
Index futures automation
- Thread starter hilmy83
- Start date
Damn, you're right. I just had bunch of things in my head and it came out wrong.
Averaging in with b/e targets is what I was referring to. I used to get into a position and aveage in; If the 1st position is a winner, i would win with the smallest position. If the trade goes against me and I end up with multiple positions, I end up trying to get out at b/e. BUt this would mean that when I actually do have to take a loss, my loss is always on the biggest position.
So my thought is to just not do breakeven and to hold on those positions to target. This would allow me to have some occasional big winners to match the big losers.
I think it works out to about 1:1 RR distribution based on average position. BUT, if market is inherently mean reverting like I mentioned, then the probablitliy of one sided move is less likely, which would indicate postiive expecatancy toward mean reverting (with averaging in).
If i add only 3 positions (equal distant), stop out on the 4th R; my reward to risk r:r for every position would probably be something like this:
1:3 (75%)
2:2 (55%)
3:1 (25%)
EV would be 2.6-2.4 = 0.2
Curently, my strategy is just 1 position, and assuming current 1 position at 60% (1:1 rr), i'm looking at 0.2 EV.
I only have about 6 months sample to support this current strategy though.
I mean it's pretty much the same. THe impact of averaging in are commissions, so technically single position makes sense. Assuming the w/r stats are correct.
1:3 (75%)
2:2 (55%)
3:1 (25%)
EV would be 2.6-2.4 = 0.2
Curently, my strategy is just 1 position, and assuming current 1 position at 60% (1:1 rr), i'm looking at 0.2 EV.
I only have about 6 months sample to support this current strategy though.
I mean it's pretty much the same. THe impact of averaging in are commissions, so technically single position makes sense. Assuming the w/r stats are correct.
Redid the math here and just based on observation over the years, I think it should be more like this:
1:3 (0.7)
2:2 (approximately 0.5)
3:1 (0.25 or below)
EV should be around -0.2. I think that as price makes a strong directional move, it's improbable that it will retest the other side of the range again. Making additional positions less likely to hit the original target. This of course depends on how wide the scale in levels are.
I have to make the assumption that after price makes a move, there will be a retracement. But if the move is extended, it will likely continue, rather than reversing to other side of the range.
Mean reversion can work with scaling in if I maintain my target and do not overcommit to scaling in a trend.
If I did just a 2nd position scale in
1:2 (0.6)
2:1 (0.5)
EV =0.3 (better than current vanilla strategy).
I have to assume my 2nd position is not going to hit stop loss most of the time and will only hit the target 50% of the time if it's within reasonable range.
I think the breakeven exit has been a crutch all this time.
I often wonder about this as well. I mean it feels good to scale into a position and then be able to exit both contracts without a loss, but sometimes it makes the loss even worse because you can't get a BE exit, and when you do get the BE exit, you didn't make any profit, just avoided a loss. In order to balance out the occasional big loser, you likely also need a big winner, not just a BE exit. But psychologically, its difficult to not take a BE exit when its presented to you, especially since it can drop again and turn into a big loss.
Last trades:
Today 30
Total 346
Not bad for $3500 for some experimental run the last couple of weeks.
I want to take some time off from the journal to figure out my strategy a little more. BRB