From theory to practice - page 594

 

Here's what I want to say to those who are suffering:

The algorithm laid out here:

Forum on Trading, Automated Trading Systems and Trading Strategy Testing

From Theory to Practice

Alexander_K, 2018.09.17 10:17

OK. Let's put it down. I do not care - just to fill my own pockets, and I do not care about other people's pockets.

1. I work with ticks in a sliding second time window.

2. for example, take a window = 14400 seconds, and create 3 (three) FIFO(14400) buffers.

3. With frequency = 1 sec. count the difference between current and previous price value (increment). Everything in a row, no matter whether it was a real tick or not, is written into buffer #1. We calculate the sum of all values in it. This is the price. Black line.

4. Count increment modules - we write them into buffer #2. Count the sum. Divide by 14400. This is the average rate of change in price. Let's call it C.

5. Now it is a little bit more difficult. We need to count the number of real ticks in this window. At every step, we look whether the increment itself or the time of value arrival has changed. If it has, we write a unit (1) into the buffer №3, if not - 0. Count the sum of the units. For example, we get 12345. This is the real number of incoming ticks in 14400 seconds. The sum of increment units from buffer #2 is divided by 12345. This is the average value of Lambda increments.

6. Calculate diffusion coefficient by formula: D^2=C*Lambda*T. Standard deviation Sigma=sqrt(C*Lambda*T).

7. Now make the assumption that all increments in BP are weakly dependent. The sum of such values gives a number belonging to a normal distribution.

6. From zero we plot support/resistance lines = +-2.5758*Sigma, where 2.5758 is the 99th quantile of the normal distribution. These are red and blue lines.

7. For the price it is the same, only +-2.5758*Sigma is not taken from 0, but from the initial reference point, i.e. the first element in the FIFO(14400) buffer.

That's it. This is the maximum that can be squeezed out of standard (not abnormal!) diffusion.


and shown graphically here:

Forum on trading, automated trading systems and strategy testing

From theory to practice

Alexander_K, 2018.09.17 09:38

If you think of the drift as a shift in the starting point, then you can also just use the price in the sliding window.

Like this:



is the definitive way to describe market price dynamics as a SB with drift and Mean reversion trading tactics.

It gives a small but positive value - about the same as the interest on a bank deposit.

I will not pay more attention to it and I suggest not to bother with it. Finita la comedia...

Finita?

No!

Let this algorithm remain basic. And I will personally look for the magic key that defines "trend/flat", which is so lacking, in neuronets.

By combining two blocks of programs: Wiener process + neural networks, I hope to behold the Grail.

Thank you for your attention.

Schrodinger's cat was with you, and A_K2 will be on the air from October. He promised.

 

Last 24 hours, 12%.

I've even emailed you how to apply the current multi-page development.

against the wall peahhhhhhh

...

 
Renat Akhtyamov:

Last 24 hours, 12%.

I've even written to you in person about how to apply the current multi-page developments.

♪ against the wall... ♪

...

But you're not giving your Grail, and A_K2 is promising all those who are suffering. And there are already 594 pages of eager ones.))

But he's already retracted his promise and said he wouldn't give it to everyone, only... ...to some who have special merit.

Now he will do NS, but in terms of trend / flat it will not give him anything. The trend/flat is visible to the eye.))

 
Yuriy Asaulenko:

But you don't give your Grail, and A_K2 promises all those who suffer. And there are already 594 pages of suffering.))

He's already retracted his promise and said he wouldn't give it to everyone, but... some who have special merit.

First they count in terms of money

before applying statistics to get an average signal

but here... ....... the other way round for some reason

 
Renat Akhtyamov:

first calculate in monetary terms

and then use statistics to average the signal.

And tutttttttt.... the other way round for some reason

The main thing is to promise. The Russian experience shows that this is the way to success. And how you count and what you get or fail to get is a matter of tenths). It is necessary later, without waiting, to start promising something else.

 
Alexander_K:

The algorithm outlined here:

and shown graphically here:

is the definitive description of market price dynamics as a SB with a demolition and Mean reversion trading tactic.

This phrase clearly shows the utter ignorance of its author.

Demolition is a continuous rise. For SB with demolition, the optimal (and only) strategy is "buy and hold". This model is usually applied to stock indices, not currencies. There is no question of mean reversion in this model.

And for a mean reversion strategy, which is usually applied to currencies, the basic model is the reversion (antipsychotic) process, with negative correlation of incomes, or (the same thing) H<0.5. But in no way an SB with demolition.

So no meaningful result has been produced by the author of the thread yet. I am surprised to see so many dilettantes who for 500 pages have been giving heed to utter rubbish but lazy to read at least the basics of describing random processes.
With such an approach there is no profit in sight).
 
secret:
This phrase clearly shows the utter ignorance of its author.

Demolition is about constant growth. For SB with demolition, the optimal (and only) strategy is "buy and hold". This model is usually used for stock indices, not currencies. There is no question of mean reversion in this model.

And for a mean reversion strategy, which is usually applied to currencies, the basic model is the reversion (antipersistence) process, with negative correlation of incomes, or (the same thing) H<0.5. But in no way an SB with demolition.

So no meaningful result has been produced by the author of the thread yet. I am surprised to see so many dilettantes who for 500 pages have been giving heed to utter nonsense but lazy to read at least basics of describing random processes.
With such an approach you will never see a profit).

It is easy to offend an artist. Name a topic where something is different.

 
danminin:

dreams... dreams...

By the time it shows a reversal, it will be too late.

When it shows a U-turn, it will indeed be a U-turn, i.e., the END of the Current Trend and the Beginning of the Next!

But will you see it or not! Watching and not SEEING is so common!!!


 
Yuriy Asaulenko:

It's easy to offend an artist. Name a topic where something is different.

Look at the old threads from 2009-2012 where random processes were discussed. Those were just monsters of their craft writing there. We are all no match for them.
 
secret:
Look at old threads from 2009-2012 where random processes were discussed. They were written by monsters in their field. We are no match for them.

And all these monsters have left both the forum and forex). The chickens are counted in the autumn. And the monsters too).

By the way, read the threads from 2008. - There people were even more monsters than in '12. And where are they all?