1st and 2nd derivatives of the MACD - page 15
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
But in the next thread the subject of boobs is being discussed and they have already started talking about soup kits...
Valera, I will tell you once again that until you understand what MACD is and how to use it in trading, its first and second derivatives have no sense. Well, not the derivative, but the difference.
I do not believe in MACD and do not even try to use it. I don't believe in it because it is too unspecific to become a profitable tool. A stupid band-pass filter, to put it crudely. What's the point of a band-pass filter in trading... I don't understand.
Alexei, a bandpass filter (preferably a selective one) is needed to decompose a complex series into a spectrum. It turns into a set of sinusoids. The sinusoid is strictly deterministic. You always know where and when it was and will be. Extrapolation and spectral synthesis gives an approximate future trend. This is the most interesting thing you can get from selective filters.
But it's not that simple :-(.
Well, good luck in the new year!
By the way, maybe you should try adding dynamic parameter calculation to your last TS? If you haven't tried it yet...
Of course, there is such a calculation. Improved the parameters a bit. Spectral analysis is strongly tied to equal-volume representation of history. Below H1 the problems begin.
I am currently writing a ticks translator from Ducas into my program. By the end of January it will be possible to run old TS on ticks.
Of course, there is such a calculation. Improved the parameters a bit.
There, you see, I'm not giving you bad advice. :)
Mathemat, the point of the picture is that the difference in readings is sometimes easier to perceive as a histogram rather than as absolute numbers. But this is subjective... Of course, you can draw a line instead of a histogram. These indicators are based on Leonid's statistical arbitrage, but apart from the basics there is nothing left. The tactics are completely different, and these indicators are used at a separate stage...
Of course, there is such a calculation. Improved the parameters a bit. The spectral analysis is strongly tied to the equal-volume representation of the history. Below H1 the problems begin.
I'm writing now a translator of ticks from Ducas into my program. By the end of January it will be possible to run the old TC on ticks.
What's the point of making ticks into equal volumes if the hell they need to be synchronised with other pairs later?
I try either to filter by H-volatility according to Shiryaev's method (besides the number of ticks, the number of inflections is analyzed), or to create a new ticks scale in the terminal - to align tick arrivals (their server time) by one general timing.
What's the point of making equal volumes from ticks if you have to synchronise them with other pairs later?
I'm trying to either use Shiryaev's method to filter by H-volatility (besides the number of ticks, the number of kinks is analyzed), or to create a new time scale in the terminal - to align tick arrivals (their server time) by one common timing.
For me, there is no problem with synchronisation. It's been solved for a long time. And I don't bother with indexes only.
What did I write about indexes?
What did I write about indices?
For me, multicurrency = indices. Otherwise, I don't understand why multicurrency?
The index decomposition is simply more compact and convenient for visual analysis and selection of the best instrument, you can also do without indices - cluster decomposition for each individual pair, it's just cumbersome, but more accurate than the index decomposition.