I don't want to hold a full size position in futures all the time. The volatility would be too much for me. I want to get in and get back to cash with a quick profit if possible. I don't have a finance background (I work in allied health care). So all the number crunching is a bit overwhelming for me.
AutoTrading Micro NQ and ES
- Thread starter rolan
- Start date
What's your average leverage applied on this (position size sum / account size)? I guess you scale it dynamically as well?
Hi Snuskpelle,
I typically keep it under 1X, but 2X to 4X is within reason for me. 100% cash going into tonight's session, I have my max position allowed set to a minimum (<1X leverage). This is the bit of discretion I allow in my system, cautiously altering the contracts traded.
Although, I was just allowed to place an order for 260 contracts of MES, so maybe I should just buy and hold.
I typically keep it under 1X, but 2X to 4X is within reason for me. 100% cash going into tonight's session, I have my max position allowed set to a minimum (<1X leverage). This is the bit of discretion I allow in my system, cautiously altering the contracts traded.
Although, I was just allowed to place an order for 260 contracts of MES, so maybe I should just buy and hold.
Last edited:
*spits beverage all over the screen.* 260 MES contracts? Your system is going to destroy you. OMFG. That is just fucking hysterical.
Folks, look at that quote as the true reason to why backtesting is just BS, and why sim is not to be done.
This is the most ludicrous idea I have seen in a future test.
STOP BACKTESTING. TRY IT IN REAL LIFE WITH REAL MONEY, AND YOU WILL KNOW RIGHT QUICK WHY BACKTESTING IS STUPID!
Here, this music might help you. Holy shit!!!
Well, the overnight session went well for me, making about $700. Then market tanked at the open. I took a 4K loss late morning. When I was aware of what happened, I revenge traded and got back in with excessive size, thinking a V bottom was imminent, which didn't pan out.
Currently down about 12K from equity peak. Now I will stop interfering and continue to allow bot to trade on its own. I will take another hefty loss as my penalty.
My goal is to trust the bot to maintain position sizing better than me, lol.
Current leverage level is 5.8X.
Currently down about 12K from equity peak. Now I will stop interfering and continue to allow bot to trade on its own. I will take another hefty loss as my penalty.
My goal is to trust the bot to maintain position sizing better than me, lol.
Current leverage level is 5.8X.
Ok, so bot took exited all earlier trades from today, including my unfortunate adds. 9K realized loss on the day. I am back in strategy mode now to return to the mission statement of this journal, of no manual long entries. It is more fun to watch trades occurring anyway than to sit on mounting loss. Hey, I would pay 9K just for a front row seat
. Lets see if I can make money while I sleep again 
.
. Lets see if I can make money while I sleep again .well done Rolan ! you are pretty balsy with the exits 
what makes you exit in the end ? thanks for starting this journal to show what can be done ! best John
what makes you exit in the end ? thanks for starting this journal to show what can be done ! best JohnHi John,
I am selling into strength, even when taking a loss. I do vary the exits though. Like after the sizable drop today, I am going for a longer hold which will still likely exit within a few hours to a day or so. Its not that I must vary the exit, I am just trying to optimize net profit.
I am selling into strength, even when taking a loss. I do vary the exits though. Like after the sizable drop today, I am going for a longer hold which will still likely exit within a few hours to a day or so. Its not that I must vary the exit, I am just trying to optimize net profit.
I did that and it confirmed that backtesting was a very good idea. Not stupid at all.
Depends of course of your ability to do a correct and realistic backtest. That's where the problem is. If you suck, your backtest will suck too.
This is true to a large extent. Another way of saying it is, everyone gets what they want out of their backtest, whether that is to show something that does not exist or show that something does not exist (note the subtle but massive difference).
If you want an overfit pile of crap for ego reasons or otherwise ignore to fix/account for realism limits on your simulation, then the results are manure. It's easy to fall into that mental trap to be fair, especially since valid genuinely potentially profitable results are hard to come by.
Last edited: