ATC without timeframe(TF) - page 2
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
Alexey, you are wrong. You yourself are working with candlesticks on 1S timeframe (not even candlesticks, it seems, but only one C price) and you say that "candlesticks are not needed".
The wheel was invented thousands of years ago by some troglodyte - and is still in widespread use.
I also didn't explain to you that any signal for representation is quantized and discretized. This is what the timeframe and candlestick representation actually does. All information inside the candlestick is not needed to work. If you do, move to a smaller timeframe.
Spell out what plz.
I don't work on M15. I don't work at all, I just watch all available timeframes. It's a robot. How do you know something about my trading - are you telepathic? ))
About sampling - according to the Kotelnikov-Nyquist-Shannon theorem the sampling frequency should be more than 2 times higher than the highest frequency in the signal spectrum. But this is for exact representation, we do not need such precision in trading.
But the sampling rate even at M1 is 1/60 == 0.016(6) Hz, which is too low. I have accepted the compromise of 1 Hz, and the processor is not so loaded and the accuracy is acceptable.
ZS: You read my post very inattentively, I wrote in Russian"The robot doesn't need any candles".
Say what plz.
But the sampling rate even on M1 is 1/60 == 0.016(6) Hz, which is too low. I've accepted the compromise of 1 Hz, and the processor is not so loaded and the accuracy is acceptable.
ZS: You read my post very inattentively, I wrote in Russian"The robot does not need any candles".
So I wrote - you work on a timeframe 1S - second. (And you're saying the robot doesn't need candlesticks but you're using one-second candlesticks...
If you do not use candlesticks, you get something like Renko bars. But I haven't seen any Expert Advisors working with these bars.
That's what I wrote - you work on 1S - second timeframe. (And you're saying your robot doesn't need candles, but you're using one-second candles...
If we abolish candles, we obtain something like reneko bars. But I haven't seen any experts working with these bars.
Ah, didn't see it, at least s made it small )) But you're confused, it's not a second candle. The candlestick assumes collecting information in 1s - OHLC, I just do a sampling of the price at 1s, analog as in CD player, only there the sampling rate is 44100 Hz.
Renko is a different story altogether. I was working with Renko for one guy, and it turned out to be a joke.
Ahh, didn't see it, at least s made it small )) But you're confused, it's not a second candle. The candlestick assumes collecting information in 1s - OHLC, I just do a sampling of prices at 1s, analog as in CD player, only there the sampling rate is 44100 Hz.
Renko is a different story altogether. I was working with Renko for one guy, and it turned out to be rubbish.
Well, that's how you get 1s candles, you just take their closing price...
I tried to get something out of reneko - it did not work, and it took much effort. So, it turns out very bad without time reference. Which means I need a timeframe... So I try to do it in seconds, or hours - it depends on the TS.
What do you mean "ticks don't have this disadvantage" ??? Ticks are just as irregular, and also have "bar skips", spoiling both quantisation and discretisation.
...
I mean, no one feeds the illusion that ticks have any "period", or "timeframe", corresponding "opening" and "closing" rates, "bars" and their "skips", "quantization". And there is nothing to talk about discretisation, they are discrete in nature. This is the only primary form of presentation of the rate data coming into the terminal, there are no others. As in auditing, you have to analyse the primary accounting documents.
Everything else, with stretches of meaningfulness, like Open and Close, or with the presence of meaning, but hiding the sequence, like for High and Low - is a way of aggregation. It is not the only one. Whoever wants, aggregates ticks the way they want. They remain the only primary, which is what distinguishes them from all other representations, traditional or newly attracted. Also, as in laboratory experiments, what is measured is logged in ink so that it cannot be 'tweaked'. Tics cannot tamper with anything, they are the original source themselves.
I mean no one feeds the illusion that ticks have any "period", or "timeframe", corresponding "opening" and "closing" rates, "bars" and their "skips", "quantisation". And there is nothing to talk about discretisation, they are discrete in nature. This is the only primary form of presentation of the rate data coming into the terminal, there are no others. As in auditing, you have to analyse the primary accounting documents.
Everything else, with stretches of meaningfulness, like Open and Close, or with the presence of meaning, but hiding the sequence, like High and Low is a way of aggregation. It is not the only one. Whoever wants, aggregates ticks the way they want. They remain the only primary, which is what distinguishes them from all other representations, traditional or newly attracted. Also, as in laboratory experiments, what is measured is logged in ink so that it cannot be 'tweaked'. Tics cannot tamper with anything, they are the original source themselves.
Powerful! There is also discretisation in time. For example coming in at an average of 2-5 ticks per second, but I do sampling every second as there is no point in tracking micro movements in forex. Yes, in terms of theory this is wrong and I get frequency distortions that would be caught by the ear when working with audio. But for forex trading specifically this is irrelevant.
In the stock market, as I know, HF trading tracks everything.
That's a powerful move! There is also time sampling. For example coming in at an average of 2-5 ticks per second, but I sample every second as there is no point in tracking micro-movements in forex. Yes, in terms of theory this is wrong and I get frequency distortions that would be caught by the ear when working with audio. But for forex trading specifically this is irrelevant.
On the exchange, as I know, HF trading tracks everything.
I'd like to know what justifies the "it's irrelevant in forex specifically" assessment.
Probably it can be no more than about some selected methods, but not about Forex in general. There is as yet no universally accepted solution, but there are highly successful approaches https://www.mql5.com/ru/forum/221552/page34#comment_6179485 which involvetaking into account and comparing tick time steps measured to millisecond accuracy.
I already mentioned https://www.mql5.com/ru/forum/259533#comment_7822778 here about the corporate effect of the microtrend, which requires exactly the sequence of ticks coming from dozens of DCs, i.e. with time steps of tenths and hundredths of a second. The microtrend gives quite real forecasts of half a point - a point and a half (0.0001); adding them to others, I've been trading for several years while arbitrage was bringing tangible results. Stopped on November 21, 2014, when the law on forex regulation in Russia passed its first reading in the State Duma (lasted a year and a half; the second, third, president's signature and publication in RG passed in a month) amid the threat of Russia being cut off from SWIFT.
P.S. Discretisation is not "there" at all, but can be applied. Whatever it takes. Both in time and space. A cleverly justified data reduction by getch (hrenfx)http://tradetrade.ru/MythBusters/2014/08/04/fundamentalnaya-oshibka-algotreyderov.htmlI mean no one feeds the illusion that ticks have any "period", or "timeframe", corresponding "opening" and "closing" rates, "bars" and their "skips", "quantisation". And there is nothing to talk about discretisation, they are discrete in nature. This is the only primary form of presentation of the rate data coming into the terminal, there are no others. As in auditing, you have to analyse the primary accounting documents.
If "primary accounting documents had to be analysed" - auditors would have nothing to do. It is their job to present the accounts in a digestible form, clearing them of unimportant information.
It is the same with quotations. In theory, any transaction, even the smallest in the market, moves the price somewhere. But in reality it is impossible to account for all of them - and there is no need to. For this purpose we have the quantization by points and discretization by timeframes. It was invented a long time ago and it is very difficult to come up with something "from above". And again, is it necessary?
I have repeatedly seen that the most sophisticated TS that take into account a lot of factors - fail/profit just like the simplest ones. That is why I think that efforts should be directed not towards inventing new TS, but towards methods of using old, well-known ones.
If it were "to analyse the primary accounting documents", the auditors would have nothing to do. It is their job to present the accounts in a digestible form, stripping them of unimportant information.
It is the same with quotations. In theory, any transaction, even the smallest in the market, moves the price somewhere. But in reality it is impossible to account for all of them - and there is no need to. For this purpose we have the quantization by points and discretization by timeframes. It was invented a long time ago and it is very difficult to come up with something "from above". And again, is it really necessary?
I've been repeatedly convinced that the most complex TSs, which take into account a bunch of factors - drain/earn just as much as the simplest ones. Therefore I think that efforts should be directed not towards inventing new TS, but towards methods of using old, well known ones.
What can I say... My advice is to read a quote from http://www.grandars.ru/student/buhgalterskiy-uchet/auditorskaya-proverka.html:
An audit is an exercise in gathering, evaluating and analysing audit evidence relating to the financial position of the economic entity to be audited.
According to the method of carrying out an audit may be:
A full-fledged audit involves a detailed examination of the entire set of primary accounting documents, analytical and synthetic accounting registers, and the content of accounting statements.
In the course of a full-fledged audit, the data of primary documents are compared with the content of analytical accounting registers (personal accounts). The compliance of analytical accounting data with synthetic accounting turnovers and balances is then established. The correctness of balances on synthetic accounts as of reporting dates is checked in the corresponding balance sheet items.
For example, in credit institutions, the following should be taken into account when carrying out a full-fledged verification of the accuracy of data reflected in the primary documents on respective personal accounts
End of quote
And about whether anything else can be invented nowadays - I have already given the examples of Brusiansky and microtrend. The right of anyone to work within the procrustean bed that others have set for him is unquestionable. San Sanych is also convinced that all structures have already been invented and there is nothing to deviate from GARCH type models. Except that the top starter was asking just the opposite, the new one.
I would like to know what justifies the assessment "it is irrelevant in forex specifically" .
It can probably be no more than a matter of methods chosen on some basis, but by no means about Forex in general. There is no universally accepted solution, but there are highly successful approaches https://www.mql5.com/ru/forum/221552/page34#comment_6179485, whichinvolve taking into account and comparing tick time steps measured to millisecond accuracy.
I already mentioned https://www.mql5.com/ru/forum/259533#comment_7822778 here about the corporate effect of the microtrend, for the catching of which we need exactly the sequence of ticks coming from tens of ACs, i.e. with time step of tenths and hundredths of a second. The microtrend gives quite real forecasts of half a point - a point and a half (0.0001); adding them to others, I've been trading for several years while arbitrage was bringing tangible results. Stopped on November 21, 2014, when the law on forex regulation in Russia passed its first reading in the State Duma (lasted a year and a half; the second, third, president's signature and publication in RG passed in a month) amid the threat of Russia being cut off from SWIFT.
P.S. Discretisation is not "there" at all, but can be applied. What is needed. Both in time and space. The data reduction of getch (hrenfx) is intelligently justifiedhttp://tradetrade.ru/MythBusters/2014/08/04/fundamentalnaya-oshibka-algotreyderov.htmlExceptionally by the fact that the order execution time for MT4/5 is at least 100ms in the most optimistic case