From theory to practice - page 229

 
Alexander Sevastyanov:

You can only make such a confident statement if you have tried all types of accounts at all 300 brokers. I write what I have tested and measured personally. And not on cent accounts at GroboForex-like brokerage houses.

that's why i'm so confident, the minimum was 20ms, to lure, then they increased it

the LMAX has a stable minimum of 50 ms on MT4

as for the mt4, the minimum was 20 mms and then they increased the stable minimum to 50 mms.

 
Andrei:
No, just stupidly starts processing the next tick while the old one is not finished for one bot on the chart....

This should not happen. If the EA's computation process has not ended on the current tick and a new tick has already arrived, then this new tick should be skipped. This is MT4/5 logic. Can you confirm your rebuttal?

 
Maxim Dmitrievsky:

I've never had a problem with it, and I've scalped a lot.

It's hard to notice in scalping when there are a lot of trades and you don't know if they are all correct...
 
Alexander Sevastyanov:

This should not happen. If the EA's computation process has not ended on the current tick and a new tick has already arrived, then this new tick should be skipped. This is MT4/5 logic. Can you confirm your refutations?

Write a simple test and see... Maybe it's just me...
 
Andrei:
Write a simple test and see... Maybe it's just me...

You are the one claiming that MT4/5 EAs don't trade the way they should. I didn't notice that, why should I write a test?

 
Alexander Sevastyanov:

You are the one claiming that MT4/5 EAs don't trade the way they should. I haven't noticed that, why should I write a test?

If you haven't noticed it and there are no problems, then it's not critical for you)) It also depends on how the EA logic is built...
 
Andrei:
This is hard to notice when scalping, when you have a lot of trades and you don't know if they are all correct...

I have never had any problems with it. On mt5 there was one time when DT changed the execution policy abruptly and positions had to be accounted for differently, and environment synchronization was slow, but then it was fixed.

 
Maxim Dmitrievsky:

that's why I'm so confident, the minimum was 20 ms, to lure, then they increased it

LMAX has a stable minimum of 50 ms on MT4

To know the minimum you have to look through all brokers and all types of accounts. My lag is significantly less.

Maxim Dmitrievsky:

There is no difference between cent and non-cent

That's the theory. But the spreads on penny accounts don't differ from one another?

Maxim Dmitrievsky:

... they don't differ at all, except for server load

If we want to check the order of loading on cent ones, we need to use more accounts and less complaints from those who lost $50, and not 5000. And the server load largely determines the speed of processing a trade order.

 
Alexander Sevastyanov:

To know the minimum you have to go through all the brokers and all the account types. I have a much smaller lag.

This is in theory. I would also say that trading conditions (for example spreads) on cent accounts are the same, right?

Well, the exact load on the cent is much higher because there are more accounts and less complaints will be those who lost $ 50, and not 5000. And the server load largely determines the speed of processing a trade order.

I'm whispering, it can not be less, show the log

for the fix protocol, the average execution is 2-10 ms (and not always), and you want to be output somewhere (not output) via MT Gateway - this is the additional costs.

at the exchange in opener ~30 ms execution via plaza

 
Alexander_K:

Hypotheses have been put forward, namely:

1. The distribution of price increments is a Student's t2-distribution.

2. The pricing process is non-markovian and NO EXACTIONS can destroy this non-markovianity.

3. The first approximation of the pricing model is a Wiener process with drift, described in terms of the Fokker-Planck equation.

Dear Alexander! I came to your huge branch where you put forward these hypotheses first. To be honest I have no time to read such a huge thread. Please tell me, what particular conclusion you came to while discussing it with other dear forum participants?

Here, you put forward hypotheses - have you checked them statistically? If yes, what optimal trading algorithms have you established from them?