From theory to practice - page 1542
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
Let's not argue...
Here are my charts for August 2019 on GBPAUD.
Show me yours. And delete your tests - you don't know how to do it, why are you showing plum TCs?
Uh-huh. That's the point - maybe Lambert will help to normalize the returnee series. The most important thing is that integral on the increments (cumulative sum) should give a stationary process with constant variance and expectation =0
I'll send my long-suffering R in the evening, and we'll see what to do... :( I'll write here :)
I don't know how Zhenya does it, he cuts the increments by the threshold somehow, I guess...
Nah, I don't trim anything! I convert the price and then take the increment with a lag (if I understand that correctly) in the period window.
And the variance on the top graph is where????
Didn't calculate it. I'm going to make some files and post them here.
How do you convert the price? incremental lag is the order of the bar from which the current bar is subtracted. If the current bar is subtracted from the previous one, then lag 1, if one bar is subtracted from the previous one, then lag 2. It is like a lag period
Here I have lag = period .
:))) I'm telling you again - show your GBPAUD charts for August 2019 so everyone can see what you did wrong.
I don't get it the way you do either, maybe it's because of working with minutes instead of ticks.
Here are the files, made with periods of 12,240,1440 minutes.
The amounts are still up to you, but I'm having some trouble with the code, it's freezing somewhere.
The only thing is how to apply this to the price.Show me.
I sent you the robot in lit. then, you can test it.
It's just that I didn't write a data dump, for pictures.
If we take the standard Brownian motion B(t) and divide it by time, the process B(t)/t would seem to be stationary (in a broad sense). It turns out that the transformation of SB to stationarity is there, but there is no use. Paradox.
PS. Just in case, let me remind you that Brownian motion is non-stationary due to growth of dispersion with time.
There is nothing to be gained on the upper chart with the MAH. At best, it is 0.
On the lower one, with constant dispersion and mathematical expectation =0, the algorithm of earnings is as follows: crossing the upper line - SELL, crossing the lower line - BUY. Exit from the trade - when it returns to 0.
This is absolutely true, and I would never have told about this algorithm if it were mine 100%.