1st and 2nd derivatives of the MACD - page 47

 
AlexeyFX:

I understand that the oscillators under the quote chart are MACDs (or bandpass filters). If so, then 2, 4, 8, ... As I understand it, 512 is the filter periods (i.e. upper or lower pass band). That is, we build MACDs with these bandwidths:

Upper Lower

2 4

4 8

8 16

16 32

....

256 512

Got it right?

 
gpwr:

I understand that the oscillators under the quote chart are MACDs (or bandpass filters). If so, then 2, 4, 8, ... As I understand it, 512 is the filter periods (i.e. upper or lower pass band). That is, we build MACDs with these bandwidths:

Upper Lower

2 4

4 8

8 16

16 32

....

256 512

Got it right?


Yes. And also 2 filters. ELF 512(like MA512) and UF 2(like close-MA2).
 
AlexeyFX:

Yes. And there are 2 more filters. A 512(like MA512) low pass filter and a 2(like close-MA2) high pass filter.


I want to understand what is shown on this graph

It is my understanding that this is the main chart on which we trade. The MACDs have disappeared here. I have this suspicion that the red line is MAKD close-64 and the blue line is MAKD 64-512. Heat?

 
gpwr:


I want to understand what is shown on this chart

It is my understanding that this is the main chart on which we trade. The MACDs have disappeared here. I have this suspicion that the red line is MAKD close-64 and the blue line is MAKD 64-512. Is it warm?


Almost. The blue one is a semblance of the MA64.
 
AlexeyFX:

Almost. The blue one is similar to MA64.


Why do you need separate MAKDs if they are not used? You could immediately calculate MAKD close-64 and MA 64 and trade on them. What parts of the grail were left undiscovered? There were three of them.

Maybe, you confirm the strength of the signal by looking at the directions of different MACDs? Suppose, if all of them point up, let us buy unconditionally and vice versa?

 
gpwr:

Why do we need separate MACDs if they are not used? We can calculate МАKD close-64 and МАKD 64 at once and trade by them. What parts of the grail have not been revealed? There were three of them.


Who could have known beforehand that this particular combination would work well in this area? Actually, this is an incorrect expression, they all work the same way, this one is just more understandable. And I haven't bothered to choose, so I just did a more or less understandable picture out of the blue in 20 seconds. You can use all 10 lines if you want. You have a loaf of bread, you can cut it lengthwise, or crosswise, into 2 parts, or into 20, whatever you prefer. The main thing is that there is nothing left over.

The third part is not disclosed at all. The first two are covered at the level of general ideas, let everyone figure out the specifics for themselves. Whoever comes up with it won't post it anywhere.

 
AlexeyFX:


The third part is not disclosed at all. The first two are covered at the level of general ideas, let everyone figure out the specifics for themselves. The first two are not disclosed at all at the level of general ideas.


No, I'm not going to post it anywhere. In general, I understand the idea. I'll work on it myself. If you decide to give me a hint as to what part 3 is about, feel free to drop me a line. I promise I won't divulge it. Thanks for the clarification.
 

if the evasions from the "true" average form a "normal" bell - what more could you want from a filter?

:)

 
AlexeyFX:



Yellow lines - opening and closing upwards. Pink lines - opening and closing downwards. The rules are as simple as rupee-twenty pants. Blue line up - catch red line down and buy, blue line down - catch red line up and sell. We hold the deal until the blue line changes direction. Blue line is horizontal - we may pips back and forth along the red one's peaks. You can use the peaks of red for scaling in. Find a dozen more entry points on your own.

Thank you. I take back what I said about promoters. I acknowledge everyone's right to trade secrets, but what is posted needs to be explained, and that hasn't been done before.

I am not persuaded by these pictures. Counter example, I won't even draw it. We take a phase with two periods. Let us enter along a high phase with a large period on a low-period phase. We draw vertical lines exactly at pivot points if both ZZs coincided or shifted if we had to wait for a small ZZ.

This is what is drawn.

The problem is that there will be a lot of false reversals and we will not be able to hold the trend. HP has the same problem. And any other filters. A new bar gives a new frequency decomposition and a corresponding number of false inputs.

And this is the problem with a method that does not take into account the statistical characteristics of the residual, which is what rules everything - the tail wags the dog. This problem is recognised by all who deny the efficiency of markets.

 
avatara:

if the evasions from the "true" average form a "normal" bell - what more could you want from a filter?

:)

And if not, the bell is not normal?)
Reason: