Currently I hired a programmer to alter my trading style into an EA.
I’ve backtested it since 2014 till current.
I accept the drawdown and it went demo since May 2015.
I’ve compete with it. Trading day by day.
Unfortunately it made more money than me by end of this week.
I plan to let it run for another week. Before playing it live with one currency pair.
Question
What do u do when a robot earns more than you. Do you let it take over? When do you go live with full monty?
It has a drawdown of 1.86% monthly worst time is around 3%. It’s monthly return is also around 1%. And this is on one currency pair.
If you have already tested the EA on a demo remember you also need to test it live. In a demo, the market conditions are perfect but in live there can be many complications. Test it with a small amount and see how it behaves. My guess is that you will probably need to tweak it a bit more after a live test.
nice stuff you have there. not much trade. turn your own trading system and made your own EA, I think that’s they way of trading. if I had such EA, I would surely keep in my stash, and put more money as divesity upon manual trade. in my opinion there’s no need to fully taking care all account with EA. any automatic system are simply based on your manual trade, and any manual trade should keep improving to survive these uncertain market :). it’s would be much better if I had 5% monthly, and 3% drawdown, sound measurable since we should paid some extra fee such VPS.
You mentioned you coded your own strategy which is great so the robot basically trades what you had it programmed to do. Not sure why you said ‘unfortunately’ it earned more than you, it is your robot so you should be happy it works they way it is supposed to work. The ‘unfortunate’ part is that you did so in a demo account and I would not really put too much into it. Run it in a small live account with $500 and see how it performs as you will not know until then.
Your Robot is doing what you programmed it to do. In other words, you are doing the job manually and automatically. There are many other factors that might affect the success of your EA when you go live. First, not all brokers would want you to use an EA. Secondly, execution speed might also be a problem when you go live. Many EAs have been tested and use successfully on our mt4 platform.
I’ll try to answer the question rather than elaborate all around
You go live with the EA after at least 3-6 month of demo consistent performance without breaking backtested max dd (not a rule, but observation)
Then you need to go live with the smallest lot possible to test the execution on different brokers. If its a scalping, you could have hard time finding a good broker for your strategy.
means when the BT shown 30%, then the demo forward test shouldn’t made bigger DD than 30. going live with small lot amount are good, but how we compare the brokers execution? is there any reviewing website for such comparison, or we should try out one broker at a times.
ECN brokers may the good choice for any scalping method, and almost brokers has these type of account nowaday.
A year ago I was in the same situation: I hired a developer to transform my trading style into an EA.
I ran it in demo environment, made about 3 to 6% per week in profit (I use leverage).
Then I had the wonderful idea to open a live account and to attach my EA and let it trade. I lost 1500€ in two market sessions. That was not good…
I decided then to make a plan and to not “trust” my EA, but to optimize carefully its parameters.
What I did with my EA was this:
[B]I let it trade demo for 3 months[/B]: during this time you see if the EA can profit from “perfect” market conditions; you also see how it reacts to news releases, what are the best and the worst days of the week and, most importantly, what hours are best to let it trade.
[B]If the W/L ratio is at least 60% and the R:R is 1:2,5 you test it live[/B]: you open a separate live account with 500$/€ in and run your EA using the parameters you analyzed during the demo test.
In this phase you optimize the parameters and rerun some tests into your live account.
Remember that a strategy that works for some time [B]does not[/B] guarantee that it will works for an indefinite time.
Always test, optimize and rerun. Always.
[I]Always test, optimize and rerun. Always. [/I]
great post dany, rarely found solid post at first try.
looking forward for your another good guidance.
welcome aboard.
Good work. I was just having a side discussion in another thread. The intraday is full of algos so building a robot became key. Mine is a “bean counter” absolute stunner. I still manually trade the longer-term trends but as professional I needed something more robust, gives me time to teach students as well as trade.
I agree with fxstrategist if you think that your EA is profitable and you have tested it enough try it also with a small amount you afford to lose and see the results.
I’d suggest 500$ and [B]nanolot trading[/B]. Your results shall be used to optimize the parameters of the EA.
Sorry for the repetitions but what I’m saying is based on my experiences.
