Frosty's auto-trading bot goes live with REAL money
- Thread starter frostengine
- Start date
Isn't interesting how the human mind works... you are mad that the BOT you created is doing well!
Just an interesting observation I think.
Kind of like when you get into a trade and lose, you want it to continue down so you can say 'thank God I exited when I did!'. And if you exit a position at a win and then it continues higher, you sit there and think why is it still going up?!?
Human psyche at it's best I guess.
yea its pretty funny...... I am just kicking myself for taking it off of live money when I did... considering i'd be in the green overall by now...... but thats how it goes.. I am infact VERY happy that it is continuining to do well again... I hope it CONTINUES to do well while I get prepared to go live again one day in the future
Are you implying that the improvements you are working on now are non-algorithm related (i.e. infrastructure, money management, fail-safe mechanisms?).
I ask this because if you are changing the algorithm, then the bot performing well now doesn't mean much for your future success.
On the other hand, if you're going to bring it back with the same algorithm, that begs the question of why to shut it off in the first place, unless you did so due to concerns about money management or the bot not having enough fail-safe mechanisms in place.
Kohanz,
The changes do not change the basic "foundation" of the algorithm.. intead most changes are dont let a position move 15 pts against before getting out.. giving the bot a few fail safe exits that are only implored in exstreme conditions..... maybe market jumps crazy after a fed meeting or something....
Other changes are filters to help predict the likelyhood this particular trade is going to be profitable before entering the trade.......
These filters do not change the basic algorithm any, it just helps to filter out the lower probabilty trades within the strategy itself..
So with that said I still want the bot to continue to do really well now because a lot of these same trades its taking now are the types I want to try and make the bot better at taking
The changes do not change the basic "foundation" of the algorithm.. intead most changes are dont let a position move 15 pts against before getting out.. giving the bot a few fail safe exits that are only implored in exstreme conditions..... maybe market jumps crazy after a fed meeting or something....
Other changes are filters to help predict the likelyhood this particular trade is going to be profitable before entering the trade.......
These filters do not change the basic algorithm any, it just helps to filter out the lower probabilty trades within the strategy itself..
So with that said I still want the bot to continue to do really well now because a lot of these same trades its taking now are the types I want to try and make the bot better at taking
Well, ANYTHING can happen...but the point is: In a prior post (about 2 weeks before the shutdown) I asked Frosty where his current drawdown stood compared to his historical drawdown average (from backtest results). I never received a response. This tells me he is either not looking at this closely, or doesn't think it is relevant.
If he was not at least 2 standard deviations above his average drawdown, then he should NOT have pulled the plug IMHO.
What we've learned here is that before committing real money to a system, the shutdown criteria must be known in advance and must be applied if it is met.
how about, taking it live after a big draw down, this seem to work for plenty of system. Of course then you have to figure out how to handle the big draw down, may be at certain amount you just put on a hedge .Thus you can have big draw down number without suffering big draw down consequence.
