1st and 2nd derivatives of the MACD - page 57

 
alsu:
You don't need to predict it, the point is that at any given time you will be sure that you entered the market at the exact price of the MA. As I recall, MA and price tend to cross. Do you want to go on with your thought?

Oops. Looks like I'm not the only one who is crazy :)

Dima: Is MA with period 20 on H1 chart and MA with period 240 on M5 chart the same thing? Or not?

Almost the same. Why don't you check it out for yourself?

I was wrong. There are discrepancies, quite serious ones, thanks to Lizaveta.
 
DmitriyN:
Is the MA with period 20 on the hourly chart and MA with period 240 on the M5 chart the same one? Or is it not?


No. They are different.

Yellow - hourly20, blue - 5minute 240 ranges. all at median prices. on the hourly chart////

 
DmitriyN:
So it turns out that the MA does not simplify the task of creating a profitable TS at all?

Of course it's easier because there was a lot less noise - in the picture you drew you can see very clearly the "angle" (or derivative in scientific terminology) at which Ma is going. When it reaches a certain value we can go out.

By using Ma we reduce the noise, but you have to pay for it with a delay. In most cases this is unacceptable, because the price has already moved away and we only open a position. To be successful in trading, we should enter the market before the main price movement begins.

 
Mathemat:

Oops. Looks like I'm not the only freak :)

I just wanted to explain ((( but apparently no one believes it without detailed calculations :`(
 
alsu: I just wanted to explain ((( but apparently no one believes it without detailed calculations :`(

The return time stats there are really nasty, that's for sure.

But I haven't really tested it myself yet. Especially on netting :)

 
Freud:

a thought popped into my head.

that if you take a fan of wands and for each individual wand build a trend line, each wand will oscillate around its trend line, but the lines themselves will have a different slope for different wands. you need to take and fit these trend lines to each other. perhaps you will get something like a common trend line with all carrier frequencies, and the frequency line density behavior (picture 2)

red - price, green - common frequency overlap.


Aha! It's like taking a bunch of mash-ups and deducing the average. The practical sense is the same. It won't help you.
 
jelizavettka: Yellow - hourly20, blue - 5min 240 ranges . all at median prices. on the hourly chart///

And don't show me the median, show me the clows... Then I'll believe you.
 
Mathemat:

The return time stats there are really nasty, that's for sure.

But I haven't really tested it myself yet. Especially on netting :)

Of course it's ugly, otherwise life would be like honey) So the most promising thing here is to work on tools with better stats... like some of my old ones.
 
alsu:
Of course it's nasty, otherwise life would seem like honey) So the most promising thing here is to work on tools with better stats... like some of my old ones.
And the fastest way to check this, imho, is to write EA. It turns out to be quite universal, just set the desired level of aggregate position.
 

Are you doing a netting version or a lock version?

And second: is the waving period fixed - or can it change according to the situation?

alsu: но видимо без подробных расчетов ни кто не верит :`(

None. Two not-so-subtle mathematicians (both traders, but inflexible nettingists) called me rude words, which made me uncomfortable.