Discussion of article "What is a trend and is the market structure based on trend or flat?" - page 3

New comment
 
Maxim Dmitrievsky:

I think it's a ph for the whole series, not for "intraday" only :) And there's a lot of drift there

de facto, intraday (on M5/M15) is a classic RandomWalk.

even the limits (sigmas) can be drawn in advance and you won't be mistaken.

 
Maxim Kuznetsov:

de facto, intraday (on M5/M15) is a classic RandomWalk

even the limits (sigmas) can be drawn in advance and you won't make a mistake.

but on some days (or periods of time) still not

 
Maxim Kuznetsov:

de facto, intraday (on M5/M15) is a classic RandomWalk

even the limits (sigmas) can be drawn in advance and you can't make a mistake.

According to my research this is on all scales, not just intraday. And if for some stocks there is a shift towards growth, for currency pairs there is usually no such shift.
 
Maxim Dmitrievsky:

and on some days (or periods of time) still not

On some days and on wandering trends, but in the aggregate no. In the market is similar. Locally there are, globally it is difficult to find.
 
Maxim Dmitrievsky:

and on some days (or periods of time) still not

in these happy moments beta! =0, and one of the few choices.

The trading task in time to determine which one is selected :-)

here is the current RandomWalk USDCHF for beta=0.


lines and limits are made very much in advance, and are unchanged.

It is repelled from the limits, for obvious reasons - with the mass of people there are volumes (all sorts of stops and so on).

 

Finally, MQL has published an article about trading! I wish there were more articles on such topics. The form has been revived at once. It's nice to see.

 
Maxim Dmitrievsky:


In other words, you have reinvented the Hurst indicator in another interpretation, which determines persistence/antipersistence of a time series, but says nothing about functional dependence (trend). And the trend itself is always there.

And how is Romanov worse than Hirst?

At least the man is working, studying something, experimenting.....

The thoughts in the article are sensible.
 
prostotrader:

How is Romanoff any worse than Hearst?

At least the man is working, studying something, experimenting.....

The thoughts in the article are sensible.

I mean the confusion about the definitions of trend and flat. It is better when there is a generally accepted definition. When a trend is taken to be something other than what it is, then something else is explored, as it were

The indicator in the article, as well as ind. Hirst and entropy ind. on discretised data shows the difference between the distribution and the SB, but says nothing about trends.

I don't mind, write what you want )
 
Maxim Dmitrievsky:

I'm referring to the confusion over the definitions of trend and flat. It is better when there is a generally accepted definition. When a trend is mistaken for something other than what it is, it's as if something else is being investigated

The indicator in the article, as well as ind. Hirst and entropy ind. on discretised data shows the difference between the distribution and the SB, but says nothing about trends.

I don't mind, write whatever you want )

To be very precise, there is no trend or flatness at all.

There is only Price(t) != Grail function, which cannot be defined in any way at all.

But the point of Maxim Romanov's experiments, as I understood, is to determine the direction of short-term price movement.

price movement, and on this you can earn a little money....

Added

Permanently earn can only when the profit can be accurately calculated.

N-р

BA against the futures on this BA

 
prostotrader:

To be very precise, there is no trend or flatness at all.

There is only Price(t) != Grail function, which cannot be defined in any way.

But the point of Maksim Romanov's experiments, as I understood, is to determine the direction of a short-term trend.

of price movement, and on this you can make some money....

Added

It is possible to earn consistently only when the profit can be accurately calculated.

Nr.

BA against the futures on this BA

to determine the price direction at a moment != to predict the next moment. Therefore it is impossible to make money on it, because this Price(t) is unknown or it is random :)

I completely agree with the second point, but for BA against futures you can use other classical tools like cointegration.

1234567891011
New comment