1st and 2nd derivatives of the MACD - page 51
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
The implication is that we analyse the market reaction to previous disturbances and predict this reaction for the future, assuming there are no new disturbances. In simple terms, "predicting the tail of the reaction". At least that's how I understood AlexeyFX's post "...the question of what will be equal to C(-1), C(-2), etc. in the absence of external influences on the market". I may have misunderstood though.
That's right.
Don't calculate filters from such an assumption, it's a temporary solution to at least calculate the filters and figure out what to do next. C(-1), C(-2)... should be unknowns, and the filters will help find them. I think so...
Suggested Junko to do the following: take a regression, in which we fit your sinusoids to kotir. the scheme is as follows:
kotir = c(1) * syn1(-1 to - n) + c(2) * syn2(-1 to - n) .....
Evaluate C(i), which will give the weights of each of your sinusoids on the right edge of the graph. Don't take many lag values from each sinusoid, mah 4 or 5, this will be determined when fitting. Forecast is a button that simply substitutes the newly obtained value for the calculation one step ahead.
All regression calculations I undertook as part of the struggle for general happiness.
In these calculations we learn a lot of useful things about your model. Junko refused halfway through.
I cited ZZ as a demonstration of the problem, and you have walked away from the problem and are advertising a smoothness in the past that you simply don't care about. The problem with tradingkng is that we can't tell the difference between a reversal and a correction. In your terminology, you can't tell on every wave is the amplitude on the right edge or the wave is still going to continue and a reversal is ahead.
The recognized problem is dispersion, which is not a constant and there are huge markets for trading this non constant. As they say, we see the speck but not the log.
Sorry, it's not us, it's you...
I can. Although I don't care at all whether it is a correction or a reversal as long as there is money to be made on the move. And smoothness in the past is essential to the task, although it's not about smoothness, per se.
Dispersion is a recognised problem of generally accepted trading systems and methods, which I don't recognise because I don't want to get generally accepted results.
Understand that this is a OTHER approach. Statistics are not used here, so the very notion of variance is not very relevant.
Sorry, not us, but you...(c)
I can. Although I don't care at all whether it is a correction or a reversal, as long as there is money to be made on the move. And smoothness in the past is essential to the task, although it's not about smoothness, per se.
Dispersion is a recognised problem of generally accepted trading systems and methods, which I don't recognise because I don't want to get generally accepted results.
Understand that this is a DIFFERENT approach. Statistics are not used here, so the very notion of variance is not very relevant.
I understand that it is customary to talk about everything here except the topic itself)))
The topic has been chewed up and worked through. Fifty pages, as it were.
Now that's a valid point. And I don't have any complicated mathematics. No derivatives and integrals, divergences, stochastic fractals and other abstruse nonsense that is very far from real life. There's an 18th degree root in one place, everything else is +,-,*,/. All the complicated stuff I've done somehow always turns out to be unnecessary. One has to understand the essence of what's going on in one's mind and that's it.
I hope you will not disappoint me by saying that the increase of degrees occurs with the increase of currencies on the same principle as described here https://forum.mql4.com/ru/19399/page18 (post MetaDriver 20.01.2012 19:51)
I hope you will not disappoint me by saying that the increase in degrees occurs with the increase in currencies along the same lines as written here https://forum.mql4.com/ru/19399/page18 (post MetaDriver 20.01.2012 19:51)
That is exactly what is happening. Only with one correction. The number of currencies to be traded does not necessarily equal the number of currencies to be calculated. More may be used in calculations.
What is the frustration?
If the author does incognitoly peek at a branch, I'd advise him to try the following construction.
Suppose we have close,MA1,MA2,MA3,MA4.... Then we build the MACD following the principle - (МА2/МА1)-1.
And then it goes without calculation of the derivative of this very MACD (each MACD separately).
Then, we can apply the same principle - (MA2/MA1)-1, but insert the MACD in this principle,
We will get something like (MACD2/MACD1)-1, actually here, maybe it will help.
i>- By the way, it would be interesting to see a further decomposition on this principle - maybe it (example in the file) is what you had in mind,
when speaking about decomposition and spectrum acceleration search?
The author of the branch probably realised that all his theories and assumptions can be checked simply by writing a script, instead of spending a lot of time checking manually and guessing.
Probably studying a lot of programming at the moment, so he's temporarily unavailable).
This is exactly how it works. Only with one correction. The number of currencies to be traded does not necessarily coincide with the number of currencies to be calculated. More may be used in the calculation.
What is the frustration?
You could probably apply the phase method to indexes already built (indexes built the old fashioned way),
you can go to each pair separately and only then go to the indices,
and you can immediately take the phase method into account when constructing the indices.