First sacred cow: "If the trend started, it will continue" - page 22

 
Magnatis писал(а) >>

This is a pattern you can spot just as well on a random ramble as on a chart of the air temperature outside your window. It's not about the instrument, the timeframe or your religion at all.

Well said of course! Nothing to add...

 
Magnatis >>: всё, что написано про тренды в этой теме – неприменимая чушь.

This is a pattern you can detect perfectly well on casual wandering as well

If you also say it works on SB, that's nonsense too.

 
Grand >>:

Хорошо сказано конечно! И добавить нечего...

Yeah, only dialectic comes to mind...

 
The only thing that is permanent here is probably the fact that the price moves over time. But how to profit from it without knowing the future point in advance is just a dream.
 

Let's assume that quotes are generated and this

generator has some parameters that describe (in the statistical sense) the current situation,

and which change from a flat to a trend or vice versa,

and according to their changed structure, we can judge about the coming flat trend;

By the way, changes in the structure of parameters serve as a signal of a possible fast change of the situation.

 

as we know any complex system, i.e. a system that is not elementary at its elementary level,

It can be described by its internal state and the state of the external environment, as applied to forex

As for forex the internal parameters may include the degree of dependence on its history, for example

Autocorrelations of currency pairs and volatility (dispersion) spectra based on autocorrelation estimates;

mutual autocorrelations of pairs and estimates of joint volatility distributions; etc.

External factors can include a set of fundamental economic indicators;


It now remains to select sets of these parameters that effectively predict

system transitions from one state to another.

 
TheVilkas >>:

предположим, что значения котировок генерируются и у этого

генератора есть некие параметры, которые описывают(в статистическом смысле) текущую ситуацию,

и которые меняются от флета к тренду и наоборот, то выявив совокупность этих параметров,

и по их изменившейся структуре,можно судить о наступающем флете-тренде;

And how do we identify this aggregate if we know next to nothing about how they are generated?

 
Mathemat >>:

И как же эту совокупность выявить, если мы практически ничего не знаем о том, как они генерируются?

there are several ways of doing this, such as factor-component analysis,

which will give us clusters of values; and cluster analysis

with extreme caution, because cluster analysis is meant to divide, which means

it will ALWAYS divide, which is what we want;

 
Mathemat >>:

Если Вы еще и будете говорить, что она на СБ работает, то это тоже чушь.

Your perception of reality is infected with a mathematical virus. Hence the bias towards simple principles. Besides, even mathematics can be approached in the right way. It is more likely to be due to some characteristics of the noise in question (yes, price movement should be considered exactly as noise).


For everyone:

Your problem is that you are trying to predict. That is, ideally you should be able to say at some point, "OK, the price will go there" and take a position in the right direction. But you will not be able to make predictions, because without some information about the banks' intentions, you are dealing with nothing more than a random movement. The message that you can determine real reversals by looking at the price (even if it's just a random movement) is stupefying: "how can we tell what's coming?!". Only it's not about looking into the future at all, it's about looking at the present moment.

 
Magnatis писал(а) >>

Your perception of reality is infected with a mathematical virus. Hence the preconceived notion of simple principles. Besides, even mathematics can be approached correctly. It is more likely to be due to some characteristics of the noise in question (yes, price movement should be considered exactly as noise).

For everyone:

Your problem is that you are trying to forecast. Ideally you should be able to say "OK, the price will go there" and take a position in the direction you want. But you will not be able to make predictions, because without some information about the banks' intentions, you are dealing with nothing more than a random movement. The message that you can determine real reversals by looking at the price (even if it's just a random movement), leaves you stumped: "how can we tell what's coming?!". Only it's not about looking into the future at all, it's about looking at the current moment.

It gets more interesting with every post. No irony!

Reason: