First sacred cow: "If the trend started, it will continue" - page 66

 
timbo писал(а) >>

Please: E[x(i)] > E[x(i-1)] for up-trend.

Agreed, with a caveat:

....where i>1 :))))

 

By the way, timbo is absolutely right when he writes E(i) - i.e. the mathematical expectation, instead of

x(i), because the concept of a trend is correct only for integral, generalized

notions, like the mathematical expectation or, if you like, the movingaverage.

 
timbo >>:

Пожалуйста: E[x(i)] > E[x(i-1)] для up-тренда.


I wonder how many of the percentage of people who read your posts knew that and therefore understood what you meant correctly
 
TheVilkas >>:

кстати, timbo совершенно прав когда пишет E(i) - то есть математическое ожидание, вместо

x(i), потому как понятие тренда корректно лишь в отношении интегральных, обобщенных

понятий, вроде математического ожидания или если угодно-скользящего среднего.


One can argue, but Matemat started the topic and probably should have given his definition, appropriate at least for this thread

 
Mischek >>:


Интересно сколько в процентах из прочитавших Ваши посты об этом знало и соответственно верно понимало то что Вы имели ввиду

You are being mocked. Haven't you noticed?

M in Soviet notation. E is in English notation.

But the Slutsky-Yule effect for smoothed average "randoms" will lead you to harmonics.

;)

And timbo doesn't see the non-random component.

I hope so, for now.

 
Mischek писал(а) >>

One can argue, but Matemat started the topic and probably should have given his definition, appropriate at least for this topic

we can, but in figuring out if the trend will continue or not will be

we're interested in some kind of robustness of our conclusions, i.e. whether it's going to change

it will change at the next random variable-quote...

What stability in a random variable? Small.

And we're interested in non-random, orderly, regularized

and that's going to be generalization, integration, summation.

 
TheVilkas >>:

можно, но нас при выяснении продолжится тренд или не продолжится будет

интересовать некая устойчивость наших выводов, то есть не изменится ли

она при следующей случайной величине-котировки...

Какая ж устойчивость в случайной величине? Малая.

А нас интересует неслучайный, упорядоченный,регуляризированный

вывод, а это непременно будет обобщение, интегрирование, суммирование.


For all the elegance of Timbo's definition of a trend, it is not the definition of a trend (imho).

But my imho will not sew up my sleeve until Mathemat gives a definition, even if controversial, but for this thread it is the only correct one.

Until then there is no sense in any answers on the branch subject since everyone is talking about his favourite thing.

And to search for a definition we need another thread.

But here and now it's up to the Mathematician to make the rules.

 
Mischek >>:


Интересно сколько в процентах из прочитавших Ваши посты об этом знало и соответственно верно понимало то что Вы имели ввиду

And I wonder what alternative definition of a trend you can come up with?

It's no surprise to me that a significant percentage of people don't know what matrix expectation is and don't understand anything at all. Therefore I repeated the same thing in different variations several times. Someone may get it and become a reason to think, to look at the problem from a different angle. Nothing will help the militant ignoramuses.

 
avatara >>:


Странная ветка - одна философия...

Скептически угрюмая.

Ни теория, ни практика не интересуют?

А Алексей?


Colossians 2.8:

"8 Take heed, brethren, lest any man lead you astray by philosophy and vain deceit, according to the tradition of men..."

 

I specified which "trends" I was talking about. My own definition of a "micro-trend" (maximum 2-3 hundred pips - for a speculator, not an investor), if anyone has read it, is synergetic (according to Svinozavr): the alignment of movements of several pairs relative to a given currency into a correlated one. But that's my definition, I'm not imposing it on others.

The question of the thread was simply to find out how others see it. There have been very few substantive answers. Almost always they were either reductions to the actual technical implementation (in the language of wigs, stochs, regressions, etc.) which didn't help to understand the causes of the trend, or outright nonsense. It's the conceptual definitions that interest me.

Here's the post.