First sacred cow: "If the trend started, it will continue" - page 41

 
Mathemat >>:

Ну все, буря, надеюсь, миновала. Теперь, надеюсь, можно и к теме ветки вернуться, т.е. к тому, о чем более-менее без отклонений говорили с самого начала?

ChachaGames и Magnatis, рекомендую вам перейти на общение через личку, т.к. то, что вы выдали здесь за последние пару суток, - офтопик для темы.

Agreed. Apologies to those who have been disturbed :)

 
newbie_d >>:

Было бы досадно упустить продолжение темы последних страниц! Если не в этой ветке, то, плиз, ChachaGames и Magnatis, создайте новую - надеюсь, не одному мне интересно. Читаю и интересные мысли приходят...

Well create a topic for "interesting thoughts" :)

 
Avals >>:

Если свести все к математике, то тренд выявляется как приращение за некоторое время по отношению к волатильности. Более частные определения тренда в ТА основываются на частных свойствах ряда цен. Полезны, но не универсальны. Поэтому тренд можно формализовать многими способами - лишь бы приносило прибыль

I agree with this thought. Breaking down price noise into trends is a method of simplifying the picture. There are a lot of simplification models that can be invented. Moreover, many people here mention such a market condition as a "flat", but it is possible to come up with a model of splitting into trends, where there will be no flat at all. Attempts to introduce different trends and different flats are a way of complicating things. And complication in the process of simplification. It is somewhat strange :) In my opinion, we should use a universal and as simple model as possible.

 

Avals >> Если свести все к математике, то тренд выявляется как приращение за некоторое время по отношению к волатильности.

Slava, you know very well that even in fluxes there are jumps that can be mistaken for the beginning of a trend. Especially in some super chaotic antipersistent wave B (according to Elliott).

Or it is possible to miss the beginning of a real trend.

What you wrote is one possible formal criteria for a trend. It is not its concept, which should go even before the criterion is defined and should always work.

 
Avals >>:

Если свести все к математике, то тренд выявляется как приращение за некоторое время по отношению к волатильности. Более частные определения тренда в ТА основываются на частных свойствах ряда цен. Полезны, но не универсальны. Поэтому тренд можно формализовать многими способами - лишь бы приносило прибыль


let's not discuss the trend and who thinks about it, let's (think) use the change in its position, and only after that can you use its direction as well. WITHOUT joking!

 
voleg2 >>:

давайте тренд не будем обсуждать, кто и как о нем думает, давайте будем (задумайтесь) использовать изменение его позиции, и только после этого вы сможете использовать и его направление. БЕЗ шуток!

The idea is right :) Although, I get the impression that only a few here are interested in finding real use for trends :)

 
voleg2 >>: давайте тренд не будем обсуждать

Oleg, create another thread, no one is stopping you from doing so. Magnatis, who thinks he knows how you trade, has already created one.

Here, in this thread, it is better to discuss the trend. In general, the market is not that big a deal, but it is a real success. But this does not mean that it cannot be understood.

I myself do not think I understand it. But at least I have proposed a concept which claims to be universal.

 
Mathemat писал(а) >>

Slava, you know very well that even in flats there are jumps that can be mistaken for the beginning of a trend. Especially in some super chaotic antipersistent wave B (according to Elliott).

Or it is also possible to miss the start of a real trend.

What you wrote is one of the possible formal criteria for a trend. It is not its concept, which should go even before the criterion is defined and should always work.

Of course it won't always work in terms of the trend continuing. Only mathematically on a very long history and if the statistical characteristics are preserved, which in practice is not feasible. Hirst, Shiryaev and Pastukhov are mathematicians and formalise it (the trend) in almost the same way. And equally useless))) I just wanted to say that useful formalizations of the trend are always partial. Particularity may be of different level - market, instrument, epoch.

 
Mathemat >>:

Олег, создайте другую тему, Вам никто не мешает это сделать. Magnatis, считающий, что знает, как Вы торгуете, уже создал.

Oleg (if that's your name), come to my thread! :) Alexei, you are mistaken - not only do I know how Oleg trades, but I do it myself :)

 
Avals >>: Херст, Ширяев, Пастухов - математики и формализуют его (тренд) практически одинаково. И одинаково бесполезно)))

I don't remember any of them having a definition of a trend. But I haven't really read them. And none of them seemed to talk about multicurrency (especially Hurst :) ).

I just wanted to say that useful trend formalisations are always private.

Of course. Any formalisation is a reduction, which is by definition private.

2 Magnatis: if there is anything of interest to me in your thread, I'll join in too.