From theory to practice - page 448
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
I feel the distribution should be the same with any sliding window, I don't know why. Maybe it's like looking at an oil painting from different distances, from far away it's clearer than from near, but the essence doesn't change.
It cannot be the same.
It has been said many times before: trading intensity varies quite consistently over the course of a day.
And the wider the window, the more the intensity fluctuates.
Also, as with a simple MA, any significant spike gives two impacts in a sliding window
- as its effect appears when it hits the window on one side and when it goes out the window on the opposite side.
Let's look at another important thing - the distribution of the sum of tick increments in the sliding window = 4 hours.
Recall that we expect to see a Gaussian normal distribution. Only then is a "return to the mean" guaranteed when going beyond confidence levels in strategies based on the assumption of a process in the market as an Ornstein-Uhlenbeck diffusion process.
Citizens gold diggers!
The Ignorance Guard Service reminds us:
a return to the mean is "guaranteed" only by the existence of a process memory, i.e. an inverse dependence of the following changes on the previous ones.
And not at all by the kind of distribution of increments.
Who does not understand this, he has an F in mathematics and perpetually empty pockets.
Good luck.
You don't need to talk), it's enough to simulate.
Generate yourself two processes with normal increments. One without memory, the second with memory (the next increment with probability > 0.5 has a sign opposite to the previous increment).
On the second your system will make money, on the first it will not.
That's all the mechanics.
You don't need to talk), it's enough to simulate.
Generate yourself two processes with normal increments. One without memory, the second with memory (the next increment with probability > 0.5 has a sign opposite to the previous increment).
On the second your system will make money, on the first it will not.
That's all the mechanics.
Why bother modelling if the price is already modelled?God, I'm tired of talking to dumb kids...
Goodbye, gentlemen!
Sincerely,
Alexander_K and Schrodinger's cat in one.
No don't go! This is the only topic I've been reading for the last six months. Don't pay attention to anyone else, go towards your goal.
Increase in incremental frequency in a 60 minute observation window
No don't go! This is the only thread I've been reading for the last six months. Don't pay attention to anyone else, go towards your goal.
Citizens of the goldfields!
The Ignorance Guard Service reminds us:
a return to the average is "guaranteed" only by the existence of a process memory, i.e. an inverse dependence of the following changes on the previous ones.
And not at all by the kind of distribution of increments.
Who does not understand this, he has an F in mathematics and perpetually empty pockets.
Good luck.
See below.
Here's the question - what time window would be obtained by processing so many ticks?
Forum on trading, automated trading systems and strategy testing
Question about arrays
Sergey Savinkin, 2018.07.13 18:35
It sayshere that the maximum size of the array is2,147,483,647 elements.See below.
There's also the question - what kind of time window will you get if you process that many ticks?
You can watch fire in the fire, water in the river and physicists trying to beat the market indefinitely)))