ATC without timeframe(TF) - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
What can I say... Advise you to read a quote from http://www.grandars.ru/student/buhgalterskiy-uchet/auditorskaya-proverka.html:
An audit is an exercise in gathering, assessing and analysing audit evidence relating to the financial position of the economic entity to be audited.
Well? Keep reading, mate!
What is the result of the audit ?
All information of the audit is contained in the audit report, in its summary. And, in fact, fits into ONE sentence - the auditor's opinion of accuracy. ALL !
A heap of accounting information at the input, one sentence at the output.
It's the same with quotes. Full quotes at the input, one bar at the output.
Exclusively because the order execution time for MT4/5 is at least 100ms in the most optimistic case
So?
The time between rate updates is, as you believe, "an average of 2-5 ticks per second". Accordingly, out of these 500-200 ms, during 100 ms the rate in the terminal differs from the server's rate, which is fraught with requotes or slippage. During the remaining 400-100 ms all is fine.
I cannot understand how you associate the time of order execution with the possibility that some of the data in the input stream may be discarded.
So?
The time between rate updates is, as you put it, "an average of 2-5 ticks per second". Accordingly, out of these 500-200 ms, during 100 ms, the rate in the terminal differs from the rate on the server, which is fraught with requotes or slippage. During the remaining 400-100 ms all is fine.
I cannot understand how you associate the execution time of an order with the possibility of throwing out some data from the input stream.
These fast changes will be filtered on the server because of the execution delay, what is not clear
These quick changes will be filtered out on the server due to execution delays, which is unclear
Are you suggesting that the flow of quotes distributed from the server to the client terminals changes depending on the appearance or absence of open/close orders, which can be delayed by execution?
What is "filtered on the server"? Where do 'quick' changes come from?
The question was about the validity of discarding some of the incoming data.
Well? Keep reading, mate!
What is the result of the audit ?
All information of the audit is contained in the audit report, in its final part. And, in fact, fits into ONE sentence - the auditor's opinion on the accuracy. ALL !
A heap of accounting information at the input, one sentence at the output.
It's the same with quotes. Full quotes at the input, one bar at the output.
Why would the result of an analysis be one bar. Usually they say one signal. And, I agree, the input is full rate data, i.e. tick data.
Are you suggesting that the flow of quotes distributed from the server to the client terminals changes depending on the appearance or absence of open/close orders which can be delayed by execution?
What is "filtered on the server"? Where do the "quick" changes come from?
The question was about the validity of discarding some of the incoming data.
I don't assume any such thing, don't make it up for me. I discard them because they are at the level of noise.
The discussion has gotten to a very narrow topic, in which many people (me for sure) do not understand much (filtering, sampling, interpolation, etc.)
If i'm not sure, then i should use timeframes as a classic strategy, or should i dig into tick strategies or ignore them?
If you have any other ideas, directions without timeframes, other than quite primitive ones (entry by time, weather, wind or even random, etc.)?
The discussion has gotten to a very narrow topic, in which many people (me for sure) do not understand much (filtering, sampling, interpolation, etc.)
If i'm not sure, then i'd better start thinking about it, and then i'll start using timeframes, because i don't know what to do with them.
Do you have any other ideas, any directions without timeframes, except for very primitive ones (entry by time, by weather, by wind, or even randomly, etc.)?
I would rather trade patties in the street.
I don't assume any such thing, don't make it up for me. I discard them because they are at the level of noise
Alexey Volchanskiy 2018.06.22 10:05 2018.06.22 09:05:18 #20 EN
I would like to know what justifies the assessment "it is insignificant in Forex" .
Only due to the fact that the execution time of an order for MT4/5 is not less than 100 ms in the most optimistic case
End of quote
And how did you manage to associate the noise level with order execution time?
The discussion has gotten to a very narrow topic, in which many people (me for sure) have little understanding (filtering, discretisation, interpolation, etc.)
If i've got a good idea, then i'll try to use tick strategies or i'll use timeframes, like i used to do classics.
Maybe there are some other ideas, directions without timing ATC, except very primitive ones (input by time, by weather, by wind or generally randomly, etc.)?
Why such restrictions, either ticks or OHLC? Why not build any graphical representation, including, above all, taking into account the nature of forex rates. There are only 7 degrees of freedom at 28 rates, that's where people who have a good idea of a four-dimensional projection of a six-dimensional cube, like Bartini, can run wild. You can analyse seven-dimensional vectors of principal components of increments.
And the "timeframe" ATS itself can be done using something else than OHLC. For example, as I said here, E1 E2 instead of OHLC. Which is close to the getch (hrenfx) ideas of reducing data at the expense of those that are known to be unprofitable.