Strange journal when data horizon is short

[Anonymous]

Hi Petko.

Great videos you make. When using EA Studio and e.g. testing H1 strategies I found a very strange thing: If I set the data horizon to only one week and look at the journal then many strategies start trading before the start date. Example:

Start date limit: 5. Feb. 2018
End date limit: 11. Feb. 2018

Journal
Id Time Type Position Amount Price Stop Loss Take Profit Profit Balance
1 31/01/2018 00.00 Open Long 1 1000 1.23946 1.23136 1.24906 0.00 0.00
2 31/01/2018 21.00 Close 1 1000 1.24148 0.00000 0.00000 1.79 101.79
3 01/02/2018 12.00 Open Short 2 -1000 1.24288 1.25098 1.23328 0.00 0.00
4 01/02/2018 13.00 Close 2 -1000 1.24469 0.00000 0.00000 -1.91 99.88
5 02/02/2018 11.00 Open Short 3 -1000 1.25092 1.25902 1.24132 0.00 0.00
6 02/02/2018 18.00 Take Profit 3 -1000 1.24132 0.00000 0.00000 9.50 109.38
7 02/02/2018 20.00 Open Short 4 -1000 1.24450 1.25260 1.23490 0.00 0.00
8 06/02/2018 15.00 Take Profit 4 -1000 1.23490 0.00000 0.00000 9.60 118.98
9 08/02/2018 08.00 Open Long 5 1000 1.22595 1.21785 1.23555 0.00 0.00
10 08/02/2018 12.00 Close 5 1000 1.22404 0.00000 0.00000 -2.01 116.97
11 09/02/2018 00.00 Open Long 6 1000 1.22448 1.21638 1.23408 0.00 0.00
12 09/02/2018 15.00 Close 6 1000 1.22403 0.00000 0.00000 -0.68 116.29
13 09/02/2018 17.00 Open Short 7 -1000 1.22762 1.23572 1.21802 0.00 0.00
14 10/02/2018 00.00 Close 7 -1000 1.22362 0.00000 0.00000 3.90 120.19

And I did hit the recalculate button :)
Same issue with M15 strategies. I haven’t tried other time frames.
Why is that?
Thanks!

Best regards

[Anonymous]

By the way: In the EA Studio Editor I noticed the following for the example above:

Data start 24/01/2018 13.00
Data end 10/02/2018 00.00

Shouldn’t these dates be the same as in the data horizon?
Thanks.

Best regards

[Anonymous]

Ups, I found out that it is because the minimum data bars is set to 300 in the data horizon settings. The field is grayed out so unfortunately I cannot change it.
Sorry for the messages. You can delete the thread if you want.

Regards

[Anonymous]

Hello Hamselv,

every thread is useful because it brings information to the others. There will be students wondering the same, so you have the answer :)

What is your idea in testing the strategies or generating them for such a short period of time? Or just experimenting?

[Anonymous]

Hi Petko.
Of course, I see. Well instead of demo trading in MT4 and waiting for that, I wanted to do it in EA Studio instead by just changing the data horizon dates. To be honest I wanted to test your M15 strategy where you replace strategies every Monday, Wednesday and Friday. So I wanted to set the date interval in EA Studio to only a couple of days to see if your strategy is in fact profitable and by how much. But the minimum date interval in EA Studio is 300 bars so that makes it difficult to do :)
Regards

[Anonymous]

Hello Hamselv,

I catch your point here, but you are missing something.

When I place for example 10 EAs from the 99(or whatever number), I choose these Experts to trade for not predefined period. I do not not how long time I will keep them on the real account.

Every other day I check if some EAs need to be removed from the portfolio, but if not, they could stay in there for longer time.

For example the Expert Advisors in the new course Top 10 USDJPY strategies, has 10 EAs that so far from the launching of the course, till the moment are all profitable. SO here is the old saying: ” When something works, do not touch it”

What I mean is that I do not trade with the EAs 2 days, I check them out every other day, but they trade longer in the account. So with a certainty you need to back test them for longer period of time.

Kind regards,

[Anonymous]

Hi Petko.

I see, thanks for the explanation. Ah yes of course, i missed that step. I just tried to include that step in EA Studio but it is pretty hard. There is no way around MT4 for demo testing I guess.

I would like to test one strategy from one of your videos. Do you have a strategy/video that gives significant results within a month? Like a proof-of-concept?

Thanks!

Kind regards

[Anonymous]

Hello Hamselv,

Glad to hear from you again.

In some of my courses I show how I go around the Demo testing.

I generate strategies from the beginning of the Historical data that I have, and for end date I use the data, one month ago.

After I select the strategies I want to use, I go over Monte Carlo and all robustness test, I include this last one month and see which strategies performed well.

It really takes a minute to do that, but it is actual testing for one month.

[Anonymous]

Hello Hamselv,

Glad to hear from you again.

In some of my courses I show how I go around the Demo testing.

I generate strategies from the beginning of the Historical data that I have, and for end date I use the data, one month ago.

After I select the strategies I want to use, I go over Monte Carlo and all robustness test, I include this last one month and see which strategies performed well.

It really takes a minute to do that, but it is actual testing for one month.

[Anonymous]

Hi again Petko.

Thanks for getting back, I understand. I have already watched some of your courses. Is there one particular course that is especially good if I want to see the way you use/describe it? And most important a course that shows realistic and significant results within a month (with or without MT4 demo testing)?
Thanks.

Kind regards

[Anonymous]

Hello Hamselv,

I show the one month testing with EA Studio in The Bitcoin Algorithmic trading in 2018 + 99 Expert Advisors.

All results in the courses are realistic, in most I prefer to use Demo accounts so the students will not take it as recommendation to trade on live account.

Also, I do not show any results of the accounts, copy trading or signals, if this is what you mean. This is totally different business, and most of the companies and people doing that are scam, so I want to stay away from this type of business. I do not want even to be associated with such.

The courses have only educational purposes and I provide the EAs so the traders can practice trading.

Let me know if you have any other questions.

Cheers

[Anonymous]

Hi Petko.

Thanks again. I am looking for a course where I can see results within a month. I will try the Bitcoin course you talk about. Thanks a lot! :-)

Kind regards

[Anonymous]

Sure, Hamselv,

The best thing to see the results is to the strategies by yourself.

There are many academies, mentors, companies that show fake results and it is easy to do so.

Thais is why I only demonstrate the process, and avoid showing trading accounts.

Kind regards,

[Anonymous]

Hi again Petko.

Thank you very much for your honesty! It is hard to find these days. As you might have figured out I’m trying to find out how good EA Studio is at generating profits before using a lot of time and energy on it. I know results vary and that you cannot make any promises. Still it would be nice to have an idea. Can you tell what range to expect please? Like in percent per year with decent risk, approximately? Is it like 10-20%, 30-50% or 50-100+% or ?? Thanks. I really appreciate it!

Kind regards

[Anonymous]

Hello Hamselv,

Again question with no answer from you :)

No such promisses, you are right.

I can tell you the following: Most of the students share with me that they get used with EA Studio and succeed to make profitable strategies after 3-4 months of using it.

It is not only about how you will generate the EAs, but how you will manage it after that.

The recent course I launched about the TOP 10 USDJJPY strategies, generated for 15 days $1000 starting with $3000, but again that is not a promise, but a demonstration.

Actually these 10 were generated with the FSB Pro.

[Anonymous]

Hi Petko.

All right I see. You give demonstrations. I get it, thanks. I have to settle with that ;-)

I just saw your TOP 10 USDJPY course because you mentioned it. It’s a very good course where you use a lot of good stuff. Regarding the section “How I have created the Expert Advisors”: You mention that you have created hundreds of strategies to come up with the top 10 strategies. How exactly did you filter out the 10 strategies from the hundreds of strategies? Are you showing this in another existing course? Did you use monte carlo/multi markets? Thanks once again! :)

Kind regards

[Anonymous]

Glad to hear from you again, and happy to hear you liked the new course.

Well, when filtering the strategies I use the same filters as I show in my other courses for EA Studio:

– min profit factor( I aim at 1.2)
– max consecutive losses
– maximum draw down

Anyway, the idea is to see strategies with stable equity line, not over-optimized.

And yes, I always filter the last strategies left into the collection with Monte Carlo(Multi Markets – sometimes I can say, not really the thing I want to rely on)

Have a great Sunday, and of course let me know if you have more questions.

Kind regards,
Petko A

[Author]

Posts