EA Forex Academy – Algorithmic & Manual Trading Courses › Forums › Forex Trading Forum › EA Studio created EAs-how long do you test in demo before going live?
- October 2, 2018 at 19:19 #6526
I must first preface this by saying that I am very impressed with EA Studio as a “newbie” trader-only been trading for 2 years part-time. The strategies that I have created are looking really good after having tested them for the month of September. I took only one EA into live trading based on the trade results in September and I also took one EA that was provided by Petko in the cryptocurrency course that I purchased live as well.
My question everyone is this…how long do you demo test the EA before choosing to go live? I was thinking of continuing to test my EAs for one month and replacing the losing ones with new ones to test each month. Is one month too long or too short in demoing the strategy? The EAs I am creating are for H1 and M15 charts.October 2, 2018 at 19:51 #6527
To be honest, there is not precise answer to that question. I am not sure if even Petko will give an exact answer. It really depends on so many things.
I have been using the FSB Pro for nearly 3 years, and EA Studio for one, and I am still looking for the golden middle when to place the EAs on live. Same when to remove them back to demo…
What I have learned from Petko’s courses( I have them all and I am waiting every month for a new one) is that strict rule should be there. For example, if you say that you will wait 2 weeks it should be 2 weeks. I did the mistake to place them earlier because of quick profits on Demo.
But I prefer more the method shown in the Forex strategy course, where Petko teaches to wait number of opened trades, and not time. Because specially the higher time frames as M30 and H1, do not open so many trades for 1-2 weeks period of time. It is just more logical to me to wait at least 10 trades to be opened and closed, to make a conclusion for the EAs. Obviously, if one-two trades are opened, even for 2 weeks, it is not enough for me.
Hope that Helps.
AndiOctober 2, 2018 at 21:38 #6530
@Andi-Thanks for responding. Yeah, I forgot to mention that I do wait for 10 trades as well. So my rule is demo for at least one month and have a minimum of 10 trades. Since I am doing this trading on the side and just starting algorithmic, I am being very conservative in using H1 and M15 charts. Once I am more comfortable, I am going to be using M5 and M30 charts to increase the volume of trades.October 2, 2018 at 22:36 #6531
It is really all about testing and testing and testing…
what I found for me to be successful is to keep generating EAs daily, testing them, and placing on live, and removing from live, and all repeats again.
However, I more and more start to trust EA Studio, and I want to remove completely the Demo testing. Petko says in his last course that the Validator does the same job, and I have convinced in that by myself, but still I want to see the EAs first showing profit on Demo and then on Live. Because he profit that we miss to take on live is too big – 10 trades in positive is something great, that I want to see on live and not on the Demo firstOctober 2, 2018 at 23:11 #6532
You are so right! Testing is truly the key to success.
I do have a process of generating EAs once a month only because I have limited resources (time and computer) to run them daily. But it also coincides with my rule on demoing first.
I actually did use the Validator tool on the EAs that I get from Petko in my courses. If the EA passes the criteria that I deemed acceptable, then I go live with it. For instance, the EA from Petko was ran through the Validator that has already been uploaded with the current historical data (I do this once a month as well). AS long as there are at least 10 trades and my other criteria is fit, then these are the only EAs that I go directly into live trading. I still put them on my demo account as well, just to keep things in sync with how I handle my automated trading.October 3, 2018 at 9:27 #6534
Very Interesting topic here, guys. Even you try to avoid the Demo, make sure always to place the EAs there, even for day or two, just to see if everything is fine with it. Check out the journal for any errors or mistakes. They are not expected with EAs from EA Studio, but…just in case.October 3, 2018 at 17:04 #6537
@Petko-You are so right. I did this very suggestion. I took one of the EAs you provided from the course, ran it through the Validator with my broker’s historical data. As long as it passes my initial criteria (like win %, equity drawdown %, # of entry indicators, etc.), I then check the journal to see if there were at least 10 trades within the last month and also whether the winning percentage was still acceptable for that month. I then put that EA on both demo and live just to keep things in sync. 🙂October 3, 2018 at 21:23 #6538
It is very important as well to see same results from the Demo account and the Live account in the broker. Some brokers do not. They provide totally different quotes on both. There could be a change as well if the spread is different.
Also, if you are trading on live account, it is good to use the Historical data from the live server(even you see no huge difference in live and Demo).
More, keep in mind that some broker show different data on the different live servers. So if you collect data from one server, and then you open another live account(the broker might automatically open the account into another server) than you will need to collect data again if there is difference.October 3, 2018 at 21:42 #6539
I have noticed that as well. I had to collect data two times for one broker…I feel they work really against the client when it comes to algo trading, with these small tricks. Also, the small Historical data when a new account is opened! Why?
It is a rhetorical question. obviously because if we have a lot data, we will have robust strategies from the start!October 3, 2018 at 21:44 #6540HaliffaParticipant
After I was scammed from the brokers, I do not belive them anymore. I am glad that Petko teaches how to avoid the scam brokers.
So when they actually change the quotes in the different servers, or provide small Historical data, it is legit, right? Even they are regulated?October 4, 2018 at 16:22 #6548
@Andi-Hmmmm….I was thinking that the small historical history was because the broker was new. Now, I’m not so sure.
@Petko-Good point. I will be sure to start collecting live data as well.
@Haliffa-My understanding is that it is legit. I haven’t found anything to say otherwise. But then again that doesn’t make it right. I can understand how the different quotes could appear from different live servers due to data transfer delay, but I’m not entirely sure this is one of the reasons.October 4, 2018 at 20:45 #6552
Unfortunately, Andi is right. All the brokers offer small Historical data for new accounts.
However, if you have a regulated broker, you will just need to collect the data. It is not that big issue, just takes time to create more robust EAs with EA Studio.
- You must be logged in to reply to this topic.