From theory to practice - page 589

New comment
 
Evgeniy Chumakov:
But somehow a strategy without stops is a bit alarming. We need to think about how to control the losses.
what if you close from the average to the edge? what's wrong with that?
 
danminin:

how is it that in the ticks view the spread is 2, but when making trades the spread is 10?

is there a slippage of 8 pips?

In the tester when the lag is 0 there is no slippage.

But if you open and immediately close a trade, there will be an average difference of about 9...11 pips(five digits), not 0...2 as you would expect from a tick chart.

Naturally the bid and ask queries do not show the real spread either.

The picture is the same on the real.

 
Natalja Romancheva:

Naturally, bid and ask requests do not show the real spread either.

The picture is the same on the real.

The spread is not important for the test, it can always be recalculated by the number of trades...

 

Andrei:
а что если от средней к краю закрывать? чем плохо?


Write a detailed algorithm which way to open from the average.

 
Natalja Romancheva:

There is no slippage in the tester when the lag is 0.

But if you open and immediately close a trade, on average there will be a difference of about 9...11 points (five digits), not 0...2 as you would expect from a tick chart.

Well I'm telling you there's slippage.

 
Evgeniy Chumakov:


Did you get anything out of it?

anything?

Yep.

 
Novaja:

I will give some current thoughts, maybe someone knowledgeable can correct: As the TF increases, we are approaching SB

What could be easier, calculate the correlation of adjacent increments on different TFs)
For SB it will be 0.
Alexander II should have done this on page 1 of this thread)
 
Novaja:
I tried it with NORM.ROB( NORM();MO;SKO) and then I check it with NORM.RASP and it does not look good, I do not know how to achieve the same beauty of distribution.
The main feature of SB is independence of increments. The type of distribution is not important.
Generate a series of increments with any RMS. And then mix them up, again with any RMS.
 
secret:
The main feature of SB is independence of the increments. Type of distribution is not important.
Generate a series of increments with any RMS. Then mix them up, again with any RMS.
Yes, thanks, will come to ACF, while I understand, that is called find 5 differences on educational pictures for children ))
 
Another interesting series of sums of increments, please make a histogram (as Alexander did).
1...582583584585586587588589590591592593594595596...1981
New comment
Reason: