Fully Automated - LIVE Trading - $50k per month

M.W. said:
All tweets deleted...
More...
I mentioned a few posts back that I didn't see the value in continuing to post live trades to Twitter. However, I am open to resuming if there is enough value for both the community and myself. I just need someone to articulate a purpose to the live trades being posted...
 
Van_der_Voort_4 said:
Since you don't use stop loss i am going to assume that all strategies are trained to take profit only.
I think range expantion is your weak spot, this is what happened in January and ended in a drowdown.Mentioned win rate is 67% from that i am interpreting that in January losing trades were larger than winners and some of them may even come from trades initiated in late December.
I would run backtest of all my strategies used in December and January across data going back as far as i could,i know you say deep learning trains your strategies,but i don't think you could come up with more than 450 strategies that concentrate on exiting trade with the profit.
You use deep learning to stay emotionally detached and of course with no stop loss capitalization is of great importance,well thought out.

I am curious if that backtest of same strategies would show any 2 months period ending in a drawdown and how often this happened and actual numbers in % terms.If each year was profitable and thats the case then it holds true that its impossible to come up with more than 450 strategies that concentrate on taking profit only,more would be the same as they are going to show increased correlation.

It is interesting that you chose 21 days and you trade futures and day has 24 hours and day trades are closed in 24 hours.i don't think you train the system on minute bars,more likely volume or tick bars are used as data input, i would say 450 strategies come from less than 5% of best scenarios.Most of them must repeat themselves over and over again logic wise-by this i mean from month to next month.Do this and that does not change,targets change all the time even if by small amounts.

In system development most strategies work until they don't.It requires deeper look into why January ended with a loss.Closed out positions with a profit and subsequent new profitable trades did not compensate for those trades that ended up closed in 21 days with a loss.They were there at all times right ,until time came to close them with a loss.

BTW Congratulations nice going and thanks for opening this thread
More...

The loss in January most closely resembles a loss from march of 2005 as far as size. Overall, the swing trading strategy has had 5 losing months. It's worth noting that I only look at the closed/realized trades when looking at a loss in January. Late January saw a lot of positions opened that ultimately contributed significantly to the 46k gain in Feb.

The day trading strategy was not yet live in January - However the backtesting results for that system shows it would have produced a gain that would have caused January to be profitable overall on a portfolio basis.
 
I remember that, was just wondering though why the existing ones got deleted...



frostengine said:
I mentioned a few posts back that I didn't see the value in continuing to post live trades to Twitter. However, I am open to resuming if there is enough value for both the community and myself. I just need someone to articulate a purpose to the live trades being posted...
More...
 
frostengine said:
The loss in January most closely resembles a loss from march of 2005 as far as size. Overall, the swing trading strategy has had 5 losing months. It's worth noting that I only look at the closed/realized trades when looking at a loss in January. Late January saw a lot of positions opened that ultimately contributed significantly to the 46k gain in Feb.

The day trading strategy was not yet live in January - However the backtesting results for that system shows it would have produced a gain that would have caused January to be profitable overall on a portfolio basis.
More...

Couple years ago i backtested strategy that has similarities to yours,it produced smooth equity curve.It had fixed R/R which forced me to recalibrate and start again instead of doing it on monthly basis.I did not have long term and short term,just one set,my number of trades was also high so this also is similar to your number of strategies.What i liked is the system spend far more time being profitable than in a drawdown and if i started with a loss it bounced back and i value what you wrote about the following month being profitable.

I would not ask to post your trades,imho all profitable trades fall under same umbrella - entry signals try to catch the trend as early as possible with consideration given to volatility band,this can be build without machine learning.

The reason you doing long term and short term which is clever is because what i mentioned in my previous post the range expantion is your weak spot and as a result you don't have trained strategies with 1:3 risk/reward.Each months that ended with a loss was a result of trades time stop 21 days and to make up for that system would require more trades with better R/R which would equal to same R/R but wider stop and larger profit,but we don't know that till after the fact (wider monthly bar) so this is useless so it comes back to trades with better R/R than those trading

If i was bored i would run backtest of previous losing months and add 10 % target to all trades,see if this results in a profitable month and if it does then run full backtest and see if there is improvement.This would improve R/R and overcome trades exited with a loss.

I read you thread just once and may not remember correctly if your day trades just have time stop and profit target.If that's the case there could be days where sharp adverse movement range expantion will continue for longer than few days and this will of course include longer term system, i would give a thought about implementing some kind of stop loss based on equity curve.

Another thing to know for yourself is R/R of profitable trades,because you don't know the logic behind them.I would call this metric "in a field R/R",gather data for executed profitable trades and calculate from their entry what was actual move against in terms of price before it hit the target-difference between drawdown of each individual trade and its profit.
This info would be beneficial to the future training used in deep learning,another criteria,no trades above certain R/R or no trades with same R/R-not to correlate too much.Above certain "in a field R/R " because these was/were trade/s stopped out on 21st day,they did not reach their target.

I am not requesting reply with any results or anything, you may post them if you wish i understand importance of IP.

Somewhere on my old laptop i had a backtest of that system with smooth equity,if i find it i will post it here,i think Win % was similar around 67 actually.
 
Last edited:
Van_der_Voort_4 said:
Couple years ago i backtested strategy that has similarities to yours,it produced smooth equity curve.It had fixed R/R which forced me to recalibrate and start again instead of doing it on monthly basis.I did not have long term and short term,just one set,my number of trades was also high so this also is similar to your number of strategies.What i liked is the system spend far more time being profitable than in a drawdown and if i started with a loss it bounced back and i value what you wrote about the following month being profitable.

I would not ask to post your trades,imho all profitable trades fall under same umbrella - entry signals try to catch the trend as early as possible with consideration given to volatility band,this can be build without machine learning.

The reason you doing long term and short term which is clever is because what i mentioned in my previous post the range expantion is your weak spot and as a result you don't have trained strategies with 1:3 risk/reward.Each months that ended with a loss was a result of trades time stop 21 days and to make up for that system would require more trades with better R/R which would equal to same R/R but wider stop and larger profit,but we don't know that till after the fact (wider monthly bar) so this is useless so it comes back to trades with better R/R than those trading

If i was bored i would run backtest of previous losing months and add 10 % target to all trades,see if this results in a profitable month and if it does then run full backtest and see if there is improvement.This would improve R/R and overcome trades exited with a loss.

I read you thread just once and may not remember correctly if your day trades just have time stop and profit target.If that's the case there could be days where sharp adverse movement range expantion will continue for longer than few days and this will of course include longer term system, i would give a thought about implementing some kind of stop loss based on equity curve.

Another thing to know for yourself is R/R of profitable trades,because you don't know the logic behind them.I would call this metric "in a field R/R",gather data for executed profitable trades and calculate from their entry what was actual move against in terms of price before it hit the target-difference between drawdown of each individual trade and its profit.
This info would be beneficial to the future training used in deep learning,another criteria,no trades above certain R/R or no trades with same R/R-not to correlate too much.Above certain "in a field R/R " because these was/were trade/s stopped out on 21st day,they did not reach their target.

I am not requesting reply with any results or anything, you may post them if you wish i understand importance of IP.

Somewhere on my old laptop i had a backtest of that system with smooth equity,if i find it i will post it here,i think Win % was similar around 67 actually.
More...

Did you run the system live and/or do you still run that system live? If not, why did you choose to not run it?
 
frostengine said:
Did you run the system live and/or do you still run that system live? If not, why did you choose to not run it?
More...

What i was backtesting is only similar to yours.

No i did not trade it live,i had a valid reason for it imho that is

I concluded this was best i could do and walked away for about year and a half,completely away from anything to do with design trading system.i burned out at that time.
 
Last edited:
felix_arb said:
Update on PnL of recent days @frostengine
More...

Was waiting until the week ended. Here are the results for this week:
Swing Trading Strategy: +$48,257 closed PL
Swing Trading Strategy (Unrealized PL): +$4,717
Day Trading Strategy: -$1,284
AAPL Daytrading Strategy: -$557

Obviously the swing trading strategy had a huge week. Closed a lot of the unrealized positions carried over from last week. The daytrading strategy had a losing week - which isn't uncommon.

The new AAPL based day trading strategy had a losing week for its first week. I will be analyzing the options trades taken by the AAPL day trading strategy to see if using shares would have been better/worse as the execution for the individual stock based day trading strategies I am still debating.

I have completed training of the SBUX strategy and will turn it live next week. I am now training an AMD strategy for next week.
 
You must log in or register to reply here.
Back
Top