Pair trading and multicurrency arbitrage. The showdown. - page 158

[Sergey Gridnev]

Maxim Dmitrievsky #:
tell me exactly what to train the neural network to do.

Profitable trading ;)

[Sergey Gridnev]

Renat Akhtyamov #:

Why the fuck would you think I don't have a balance?

;)))

Then why are you giving wrong advice?

[Maxim Dmitrievsky]

Sergey Gridnev #:
Profitable trading ;)

It's purely my profile to make her learn to trade profitably :)

[Renat Akhtyamov]

Sergey Gridnev #:
Then why are you giving the wrong advice?

Why-- wrong.

Learn the ropes.

Look at the yen.

It's 100 times bigger than it needs to be????

put GBPAUD on a triangle and name what trades and how much volume you need to open, so that the equity does not move.

open a triangle on demo, show the beginning and the result in a week or two.

Can you do it?

If you can't, which I'm absolutely sure you can, shut up.

[Roman Shiredchenko]

Maybe someone will write a certain set of conditions and their variants in a branch on the current page for input and output and others.....

What are we coding?

What we are working on now......

[Maxim Kuznetsov]

just practice...

из https://www.mql5.com/ru/forum/448777/page16#comment_47761562

remove all averaging, locks, inter-locks (since multicurrency) and martingales.

otherwise it's self-deception

and only the thoroughly cleaned algorithm is tested (not optimised!!) to make sure that it does what it is supposed to do.

Your "1-2-3 years of trading and then losing" probably don't say anything about the algorithm. It is harder to get rid of martin than to make it

[Sergey Gridnev]

Renat Akhtyamov #:

from what ... wrong.

study up

Look at the yen.

It's 100 times bigger than it needs to be????

Once again:

1 lot for eurusd = 100000 eur

1 lot of eurjpy = 100000 eur.

This means that in the triangle eur-usd-jpy the volumes of eurusd and eurjpy should be the same (but differently directed).

You write "divide by the pip value". So you divide it, get an imbalance and wait for it to converge. And when it doesn't, you write: "They broke my indicator!".

[Renat Akhtyamov]

Sergey Gridnev #:

One more time:

1 lot of eurusd = 100000 eur

1 lot of eurjpy = 100000 eur

This means that in the eur-usd-jpy triangle the volumes of eurusd and eurjpy should be the same (but differently directed).

You write "divide by the pip value". So you divide it, get an imbalance and wait for it to converge. And when it doesn't, you write: "They broke my indicator!".

for the second time.

matches and potatoes have doubled.

i.e.

matches from 1 rouble to 2 roubles.

potatoes from 10 to 20.

Think and don't mess with my head.

[Roman Shiredchenko]

Maxim Kuznetsov #:

just practice...

из https://www.mql5.com/ru/forum/448777/page16#comment_47761562

remove all averaging, locks, inter-locks (since multicurrency) and martingales.

otherwise it's self-delusion

and only the thoroughly cleaned algorithm is exclusively tested (not optimised!!!) to ensure that it does what it was designed to do.

Your "1-2-3 years of trading and then losing" probably don't say anything about the algorithm. It is harder to get rid of martin than to make it

О! Thanks. I'd forgotten about that.

Yeah. We need to weed out all martins in any form...

Maybe a few smeared entries on the same volume up to three times.

Like in the delta range of 30 to 60 pips. In the direction of the pattern.

[Maxim Kuznetsov]

Sergey Gridnev #:

One more time:

1 lot of eurusd = 100000 eur

1 lot of eurjpy = 100000 eur

This means that in the eur-usd-jpy triangle the volumes of eurusd and eurjpy should be the same (but differently directed).

You write "divide by the pip value". So you divide it, get an imbalance and wait for it to converge. And when it doesn't, you write: "They broke my indicator!".

and it's also fun to count the margin :-) for some reason many people forget that trading is marginal, in debt with leverage of 1:100, for example.

and 1 lot of EUR at different times costs different amount of USD (which is usually a cutlet).

and that the effectiveness of the strategy in the end is how successfully the borrowed capital (which is 1:leverage) is used.