- This topic has 29 replies, 8 voices, and was last updated 2 years, 9 months ago by
.
-
Posts
-
-
I thought I would show what I have been working on to fully automate, analyse, filter and copy trades to live account. I have been working on this system for 12 months. I hired a Russian programmer to help me create the system. It has been a big mission costing me a fair amount of $$$ and probably a 1000+ hours of my time testing and debugging etc….
Some people may find it interesting, others may see it as an overkill but it works for me….. I show it here to show what can be done if you want to go down this path and to share my findings. I am not looking to sell the system, or certainly not at this stage anyway….
There was a few people on the other forum that talked about doing this and also there was the MT4 Tracker that someone created that was good. Any way I got sick of waiting and decided to create exactly what I wanted that suited my trading style.
A lot of the inputs / settings are based on what Petko teaches here. Basically I wanted something that would track all info from the demo, allow me to filter it in as many ways as possible and then automatically send the selected trades to a live account.
I will add some screen shots with explanations:
I run a tracker EA in the demo accounts which captures all the account data and sends it to a Mysql database. The database is capable of handling 1000’s of lines of data. I have had as many as 12000 strategies and 100’s of accounts running.
The database holds all the data and carries out some of the equations:
-
Then I use a windows based app to carry out the filtering:
I have 8 different live accounts assigned to the app. I assign the magic numbers to the accounts. the magic numbers are critical. I use what Petko teaches here to create the magic number, but the middle 3 numbers are the batch number that allows me to group certain ea’s together.
I can filter the trades from the demos with any of the filters, I can set using days and trade no. I have 3 different time / trade periods and can also select top 10,20 etc….
-
-
-
-
-
Hey Richard,
This is very interesting. Please keep us informed, not only on the development, but the results of you “live” demo account.
The question I have for you is, if you choose to change the criteria by which the system selects the EAs to move from demo to live, how easy is it to do?
Cheers,
Ilan
-
The settings for each of the filters can be changed instantly. If after the change there are strats that don’t meet the new criteria they will be removed immediately from the live account…. And any strats that now meet the new criteria will open on the live the next time they open in demo.
I watched your course the other day and have started a few trials running that will be ready in 2 weeks. I am thinking about the best way to automatically send the chosen strats to the live. Either by TOP # of each pair or other ways. Anyway all these options can be selected so that it will run on auto….
Image of pop up box to change settings instantly attached. This is from your course that I started a few days ago. Took a beating overnight but still looking good.
I like your concept of changing out weekly. Is there a reason you don’t close down on Friday end of market and avoid potential gaps over the weekend instead of Monday morning…..
-
Hey Richard,
I am very happy to see when someone does something extra and useful. Usually, most traders wait and depend on the others…so good job to you!
I am still not sure if I understand how this app works and if a database is actually needed, but I am super happy that my courses helped you out!
Keep us updated with your work!
-
-
-
-
-
-
-
-
-
-
Hey Petko, regarding the Database. It basically does a similar job as FX Blue but automatically and Live. It will show all of the strats in the app. Then you use the app to filter out the ones that you want to trade. Example: Allow the first 4 trades to be copied to the live account but then only copy trades that have a 1.2 PF over the last week and also a net balance of $1 and a R2 above 50% over the last month, or whatever you want. I can use any of the filters in the images above in any sequence and send only the selected ones to the live account. Basically exactly what you do manually when you select which strats to take to the live account the system will do it all Live and Automatically….
-
-
Thought I would post some results from the last week:
This is from Petko’s Top 30 for October. I missed the first week of the month as I had a bug I had to sort out in the system. So this has just the completed weeks data at this stage.
I am running 3 side by side tests and comparing the results:
Test 1 – No filters, so basically the demo trading and all trades included. I like to use this as my base and then see how effective the filtering is at increasing profitability, reducing DD etc….
Test 2 – Strict filter applied. Do not copy the first trade. Apply a 1.2 PF to ALL trades and also a 1.2 PF to the last 4 trades. This means all strategies need to perform well from the beginning and continue to perform or they will be dropped.
Test 3 – Relaxed filter applied. Allow the first 3 trades through with no filtering. Apply a 1.2 PF to the last 4 trades. This allows the strategies to get underway for a few trades without being removed too early and the 1.2 PF to the last 4 trades ensures that they are then performing reasonable well. Strategies that get off to a poor start but then improve and meet the 1.2PF for the last 4 trades criteria will then be selected
-
-
-
Test 3:
This is more what I am looking for…. Nice consistent gains. Another thing I am just starting to focus on is how well a Test handles things when the market moves against your positions. You can see in the demo that that big drop midweek was handled well by test 3 with a loss only about half the size of the demo. This is exactly what I am hoping to see….
You can see test 3 is now only running 20 of the 30 strategies that the demo is running. So 10 have been filtered out as not performing at this stage. Test 2 the stricter filtering has even less.
So for me finding the right filtering balance is very important. What I have learnt over the last few months is that light filtering works better then heavy filtering. If I try to remove too many strategies with strict criteria I don’t get as good results….
Light filtering appears to give the strategies enough room to prove themselves without getting taken out too early…
Any way these are just my ideas so please don’t take them as anything more than that….and test and prove your own systems….
-
This is the result from the first week of IIan’s Weekly Robots Course. (I’m hoping it is OK to post these results….. Let me know if not……)
There is no filtering applied to these strats as they are already filtered for the week before installed. So they are simply left to run for the week as is. I definitely like this approach and have started running some demo’s looking to set up a similar system myself…..
But as you can see, a very successful week……
-
-
On Petko’s top 30 above it is set to live. The tick in the tick box next to the account number means that the filter is applied automatically and is live. So as trades close in the demo, the filter fields on the app are updated instantly. If those closed trades mean the strategy now meets the filter criteria that I set, then that magic number will be sent to the live account and held there. The next time the strategy opens / closes on the demo it will be auto copied to the live. Same as if a recently closed trade puts the strategy below the filter criteria the magic number will be removed from the live account. This way everything happens automatically and in real time…
I can also momentarily check the tick box and then untick it. This will send the filtered magic numbers to the live account but because it is then unticked the magic numbers will stay the same and not update as described above. This is what I am thinking I will do with my new demos when they have run for 2 weeks to emulate what you are doing. So i will set weekly and monthly to 1.4 PF, min of 3 trades and each week send the filtered magic numbers to the live. This will update the Live account. If any of the strats are the same it should just keep those running, remove any that no longer meet the criteria and then start opening in sync with the demo. The only issue I see with this is any strats that are already running on demo won’t start straight away on the live. I will have to wait for the next time they open on demo before they are copied to live. Anyway I will run some side by side tests with my system automatically running and manually placing them as you currently do and see what the results are….. Hope that makes sense……
-
So this is how I would set up to trade what you do.
Minimum 3 trades. So you can see only strats with more than 3 trades are displayed. Then 20 days and 5 days set to 1.4PF, My system only reads trading days so this is 4 weeks and 1 week.
As you can see these strats were only started a few days ago so not ready yet but the system will automatically capture the last 5 days and 20 days and calculate the PF etc. Those that meet the criteria will stay displayed.
So that i send an even number of strats from each of the 4 currency pairs being traded I have also selected that only the top 3 of each currency are to be sent to live. (See pop up box) You can see this by the green boxes next to the strats. Again if a recently closed demo trade is better then 1 of the green strats then it will push it out of its spot and replace it……
-
-
-
- You must be logged in to reply to this topic.