From theory to practice - page 272
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
Me and Novaja have done a little research.
The tics come into the terminal at some random intervals. Alexander wrote that it's important to know what kind of distribution it is. I took the tick history from mt5, so I can work with millisecond accuracy.
...
The formula is:
gamma(4e+00, 2.8e+01) * 5e-01 + cauchy(0e+00, 7.37e+02) * 2e+00 * 5e-01
This is a formula for a specific dealing, all have slightly different parameters and coefficients.
In general, the function looks like this
function(k, Θ, γ, c){
gamma(k, Θ) * c + cauchy(x0=0, γ) * 2 * (1-c)
}
parameter c is from 0 to 1
Please tell me, @Dr. Trader, among the DCs that have slightly different parameters and coefficients in this general formula, have you come across any that quote not 5, but 4 digits?
Gentlemen!!!!!
Knowing that it's in the bag and my pockets are ready, my left eye is starting to twitch from the unbridled thirst for money... :))))))))
Even on the first pages of the thread we found out that a few hours)))
To all: Of course every brokerage company may have its own quote filters. I don't think it affects anything with such durations of deals) Am I the only one who understands it?)
Even on the first pages of the thread we found out that a few hours)))
To all: Of course every brokerage company may have its own quote filters. With such transaction durations it doesn't affect anything) Am I the only one who understands that?)
Of course not, see https://www.mql5.com/ru/forum/221552/page179#comment_6499483
Gentlemen!!!!!
Knowing that it's in the bag and my pockets are ready, my left eye is starting to twitch from the unbridled thirst for money... :))))))))
Even with the theory ready, the transition to practice is not easy. Believe me.....
Gentlemen!!!!!
Knowing that it's in the bag and my pockets are ready, my left eye is starting to twitch from the unbridled thirst for money... :))))))))
It's not hard to do in principle. In that graph, the Gamma distribution is already almost zero after 400 ms.
Roughly speaking, anything longer than 400 ms is already Cauchy.
But values with a pause < 400 ms cannot all be left. A random number between 0 and 1 (uniform distribution) can be generated. If this number is less than [koshi / (gamma + koshi)] from the formula, we also discard this tick.
I generally like this idea - if a tick doesn't come for longer than 400ms then something is wrong, and a tick that finally comes will be "bad". Better to wait a little longer for the next one.
No, Doc. That's where you're a little wrong.
One more time. Here's what you, along with the most adorable Novaja, have found:
The formula:
gamma(4e+00, 2.8e+01) * 5e-01 + cauchy(0e+00, 7.37e+02) * 2e+00 * 5e-01
This is a formula for a particular dealing, all have slightly different parameters and coefficients.
In general the function looks like this:
function(k, Θ, γ, c){
gamma(k, Θ) * c + cauchy(x0=0, γ) * 2 * (1-c)
}
the parameter c is from 0 to 1
Jesus! Well, you're going around and around!
One has to PREVENT thegamma(k, Θ) * c flowgiven by the generator of this distribution. PREVIOUSLY! Reading both real ticks and pseudo ticks. ALL.
You'll get the swellest BP increments and trading intensity in the sliding window.
Well, I don't know how else to explain it...
Tell me please, @Dr. Trader, among brokerage companies that have slightly different parameters and coefficients in this general formula, are there any that quote not 5, but 4 digits?
No, I only studied five digits.
A couple of regular dealing centres I checked had K=4, no less.
I also tried MetaQuotes-Demo - K=1.38 there, but if you round K to 1 or 2 the distributions do not coincide very well. But the noise in tick times is different there as well, the "fence" with a non-integer period.
Also there was an interesting result on ticks downloaded from the broker's site, everything less than 500 ms is very thinned there. Plus again the "fence". If we try to average it all and describe by a formula, K=200.
It seems to me that all public sources of ticks - somehow distort the time, add plus or minus 10 milliseconds or round it off.
It is strange that dealing companies do it, because their order execution time is in hundreds of milliseconds, no one would trade on ticks anyway.
One has to PREVENT thegamma(k, Θ) * c flowgiven by the generator of this distribution. PREVIOUSLY! Reading both real ticks and pseudo ticks. ALL.
You'll get the swellest BP increments and trading intensity in the sliding window.
Well, I don't know how else to explain it...
I think I got it.
Step 1 - Make your own clock generator, with random pauses between each clock, like that distribution. At each cycle, take current bid/ask price and make time series of them.
Step 2 - I don't understand it yet, I need to think about it. I won't be able to trade such time series, it's already HFT.
Am I correct in assuming that the K-factor of the Erlang distribution is used to judge the fairness of the dealing?
A bit of specific humour: