Machine learning in trading: theory, models, practice and algo-trading - page 271
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
What kind of "genius" measures price dispersion? Of course we're talking about returns or log returns.
) And what's the problem with calculating the variance of the price-the average of the time series can you calculate?
Not a problem, but not very meaningful, knowing that the process is aggregate. Just as we do not really care about the absolute value of the price, but only its change in the next moment. We have to become skilled in predicting the changes, and what kind of trajectory they will draw in the sum is secondary.
Not being stationary does not mean not being predictable
I probably didn't put it that way, it's my fault ...
By stationarity you probably mean something like - we have a price, it is not stationary, we differentiated it and it is already stationary. I mean something closer to fractality, which is non-stationarity for me. I believe in patterns no matter how ridiculous they are, but I also understand that they are fractal and that they never repeat exactly as they did in the past.
The main problem is the rapid reaction of the market to the new information.
I agree completely, even self-tuning systems are not competitive here
which is not deterministic in any way
I haven't given up yet, even the same levels, we think that the price burst for no reason, but it rebounded, we just don't know it and don't understand it
Why do you need stochastics?
I'm making fun of him, you don't get it? :)
But indicators are not only smoothing, like "levels", could be one of the most important signs, I mean the levels that we see on the chart with our eyes, where people put stops. I mean those levels that we see on the chart with our eyes where people place stops.
i've tried, am trying and will continue to try, levels are one of the strongest features the market has, in addition the level is stationary
Not a problem, but not very meaningful, knowing that the process is aggregate. Just as we do not really care about the absolute value of the price, but only its change in the next moment. We have to become skilled at predicting changes, and what trajectory they will draw in total is secondary.
"aggregate process", "returns" .....
That's it, you've taken the price increments - the series of increments is stationary. What's next?
"aggregate process", "returns"....
That's it, you've taken the price increments - the series of increments is stationary. What's next?
First, you need to check the hypothesis of (non)linear dependence of the future return(Rt+1), on N past(Rt,Rt-1,...,Rt-n) of this instrument and other liquid instruments.
To begin with, it is necessary to test the hypothesis of (non)linear dependence of the future return(Rt+1), on N past(Rt,Rt-1,...,Rt-n) of this instrument and other liquid instruments.
What if a series of price increments is "white noise"?
"White noise" is stationary...... Ah, "returnee"?
What if a series of price increases is "white noise"?
It is not.
Why?