- This topic has 179 replies, 27 voices, and was last updated 1 year, 8 months ago by
.
-
Posts
-
Edu, what Petko does in this videos is to show the updates in the functions of EA Studio.
Each one of us should fine the most strict criteria according to the Historical data each one has.
If he shows the stricter he uses(he has a lot of data) and we test with such criteria, we would not see any strategies.
Actually the idea here is to test the strategy with not all of the data. OOS helps to test the EAs as they were on Demo.
Also, when we optimize the EA, it is good to use the OOS, because this way we will see if it was over-optimized.
Now with the new updates in EA Studio, we can filter the results very well by setting the criteria, but for sure, use stricter criteria as much as possible.
Hello Petko,
I used this method with 250 trade, generated 10 EAs in total for 10 different currency pairs and placed on Demo. The first trades are all on profit.
Will see ow it goes. Thanks a lot for the awesome videos you are doing!
You are welcome, Chingi!
If you use 10 different currency pairs, you should make sure that you place equal exposure to the different currencies.
Do not place only USD pairs for example, because this might bring huge risk to your portfolio if there are some unexpected moves with the USD.
Hope that makes sense.
Hello Petko,
today i was creating some EAs and i generated this. It seems too strange to be good even if there are a lot of orders opened (more than 300). Shouldn’t this method generate only not over-optimized strategies? (just a question)
<script async src=”//s.imgur.com/min/embed.js” charset=”utf-8″></script>
Wow, Edu! How did you come up with such an equity line?
Did you use optimization in In Sample part only?
Also, how did you generate the strategy?
Hi Petko
I used the exact procedure you used in your video.
I imposted out of sample 30% and r-squared in optimization.
I used the reactor as you teach in the video.
Hello Edu,
really hard to say just from looking at the picture. It looks really too good to be true.
If you want me to have a deeper look at the EA, you can drop it in the e-mail to Marin, and I will have a look at it during the weekend.
Cheers,
Yes, it looks like over-optimized strategy. However, the idea of these tools as Monte Carlo, and OOS, and Multi Markets is to recognize these strategies and eliminate them.
It happened again but this time on another cross. The others strategies look normal but this 8000 trades? What??
I did again the montecarlo test and it resulted over optimized
<script async src=”//s.imgur.com/min/embed.js” charset=”utf-8″></script>
Jacpin, is right, I was wondering as well how you succeeded with this result. Did you get few of these strategies or just one?
If you got few in the way I showed, I would suggest you to drop them in on Demo account and see what happens, even you saw the bad results in Monte Carlo. If it is just one, than I would say it it some good work of EA Studio to find such a “fantastic” strategy.
Yes only the two i posted. The others are good and “normal”. This week i will the test them and i will keep you updated when i have 10 orders closed
- You must be logged in to reply to this topic.