Which mash-ups are the best to cross? - page 5

New comment
 
sergeyas:
MACD also expresses discontent....
and the MACD is probably the shittiest gauge - only if you look out for wiggles
 
IgorM:

http://www.priceactionfx.ru/2010/12/blog-post_15.html

trading on Monday night gaps, even at Forex, is very effective - I checked it, one thing is bad - the brokers who are open on GEP widen the spread, and not every Monday there is a tangible price jump - sometimes you spend a couple of weeks "working at the computer till night", but there is no effect ((

Bullshit. Authoritative - rubbish! ))) Doesn't it convince you? Never mind - it doesn't really matter...))) // I'll bring you some statistics... Oh, wait for it...

Shit! Purely for the sake of common sense (a rare beast!). What does the intraday trading on Stt. have to do with what you see on the same Std. In general, on Fridays at 11 piem. finish. Ahh... I've written before. Suit yourself...

It's relevant, though. It has the same effect as one sample to another. But then it is not minutes and not time-framing at all. But here it is about others. And in the context of this Dr. - Gaps have nothing to do with what you do on a shallow TF, namely, filtering and looking for crosses of sluggish schizophrenia with raving insanity, slightly smoothed by relanium...

 
Jingo:
And the MACD is probably the shittiest gauge - only if you look out for vortexes
I remember developing a superindicator... ran it, tested it, then did another one... now I look and admire... all of them are almost identical to MACD. And different periods, different methods. And the coincidence is 99%. MACD is the best oscillator))
 
zfs:
I remember developing a superindicator... running it, testing it, then another one... now I look at it and admire it... all of them are almost identical to MACD. And different periods, different methods. And the coincidence is 99%. MACD is the best oscillator))
It depends on what you see) for my tests with out of sample it's a really piss poor measure, whether it's purely a mcd or neurosti) - I don't think there is a robust system with a mcd at the core
 

I can assert that using only macdi you can get something slightly profitable.

I've also made profits with neural networks, but the optimization is a big problem, because it takes about an hour to optimize 3 kt already.

 
zfs:

I can argue that using only macdi you can get something slightly profitable.

I've done profits on neural networks too, except there's a big problem with optimization, because it takes about an hour to optimize 3 kts already.

And this "slightly profitable" as long-lived without interfering with new optimisation?

That's the main root of the wretchedness of standard macdies and their combinations.

Unstable in time - hence its in the museum.

 
zfs:

I can argue that using only macdies you can get something slightly profitable.

I also profited on neural networks, except that optimization is a big problem, because optimization of 3 kt already takes about an hour.

And if you use non-integer periods and a shift... m-muh! - years! But what a result!

Laughed out loud! I mean, laugh out loud...

 
Svinozavr:

And if you use non-integer periods and a shift... m-muh! - years! But what a result!

Laughed out loud! I mean, laugh out loud...

Haven't tried non-integer periods, although there is something to that. Recently tried to stabilise the averaging period to make the system more static. It had an effect in principle, but subsequently gave up on the idea. So you can keep on laughing))
 

in general, everyone looks at the whole berry bush from his or her own perspective)

observations take on a complex, abstract and sometimes inexplicable form.)

 
Jingo:

in general, everyone looks at the whole berry bush from his or her own perspective)

observations take on a complex, abstract and sometimes inexplicable form))

I would like to share one more observation: the method of brainstorming in us smoothly (and often!) flows into the method of free association. And here it is time to invite specialists of another profile!)
123456789
New comment