Euro Rat - page 9

 
elmucon:

and a short video simulating real trading

one month (because of very long testing)

But I assure you - in other months trades the same ...


How quickly from a drawdown gets out, just a couple of days and all, mine works on the approximate principle, but in a drawdown on a pair can sit up to two weeks, but also the drawdown of course less, but longer

 
elibrarius:
Well, the Expert Advisor is ready on full automatic - you can test - just add the forward option in the tester. During the weekend, you can probably carry out all 12 tests using the method of rolling forward.
For example I was worried about 70% drawdown in your back-test, and where 70% (in known area) - there and 100% (in unknown area) can easily happen. Forward can show that.

and I'm not embarrassed by anything ... no one is stopping me from closing the loos if the situation looks critical...

and the risk is excessive - bet half as much and the drawdown will be less ... 150 000 will do with 5 000 there is a difference of taste and colour

if you want to make a profit, you need to take a risk ... )))) or you can be like a snail slowly but surely ...



doing a test for you in 2016 - I worry like a child ...))

 
elmucon:
doing a test for you for 2016 - worrying like a child ...)))

Only do it with a forward. Otherwise there is no point

 
elibrarius:

Only do it with a forward. Otherwise there is no point


Well, let's be very frank -

With those settings that the optimizer gave the robot leaked in April (but for 4 months it has really risen)

But I put the minimum settings - by eye - without optimisation, that is, as I would have done when installing it on a real account with a large depot

uy


And as you can see - by 2016 the robot passed with two "stumbles" ...


123


let's see the drawdown ...


321


oh - 34% and the income? doubled over the year ...

But since I'm not interested in waiting a year I'll take my chances and keep an eye on it ...

 
elmucon:

Well, let's be very frank -

With those settings that the optimizer gave the robot leaked in April (but for 4 months it has really risen)

But I put the minimum settings - by eye - without optimisation, that is, as I would have done when installing it on a real account with a large depot

And as you can see - by 2016 the robot passed with two "stumbles" ...


let's see the drawdown ...

oh - 34% and the income? doubled over the year ...

But since I'm not interested in waiting for a year I'll take my chances and look and push buttons with my hands ...

What about optimizing with forward? Maybe then something better will come out?

 
elibrarius:

And optimize with a forward? Maybe then something better will come out?

Maybe the pairs should be chosen differently, abandon dollar pairs and take crosses only

 
elibrarius:

How about optimizing with forward? Maybe then something better will come out?

Although it might be better to optimise by valving forward. It might be more stable...

 
Vitaly Muzichenko:

Maybe pairs should be chosen differently, abandon the dollar ones, and take only crosses


you're right about the pairs - there are scripts to determine the smallest correlations -

I'm just saying before the spreads were too tight - now you can try other pairs...

 
elibrarius:

Although it might be better to optimise by valving forward. Maybe it will be more stable...


you understand - i've already spent a lot of time on all sorts of tests

Take my word for it, the robot will work on any part of history....

You may try to predict all the nuances of the market, but you cannot.

i think that the main problem has been solved - a robot earns with a minimum number of input parameters

There is practically no need to change these parameters and adapt them to different parts of the history...

The only question is where to look to - I showed you where to look to.

the only question is where to look - I showed you where you can look, and the tests only prove my point ...

i am not trying to make a grail - i am making myself a "cheat" to play forex ...

now i will play with cheats ...

 
elmucon:

But what a person sees with his eyes and what he can think about and make a decision cannot be programmed.

I agree, that's why I am not trying to programme robots now, but to trade manually. But for that I need time to gain experience, i.e. to lose more than one depo.

Reason: