From theory to practice - page 529

 
Yuriy Asaulenko:

That's not my suggestion).

Pardon me, it was suggested by secret.

But the question remains open nonetheless.

How to define 'only a flat'.

 
Cat Libre Black:

Pardon, it was suggested by secret.

But nevertheless, the question remains open.

How to define "only a flat".

Either on history, or not at all).

 
Yuriy Asaulenko:

Either on history or not at all).

Agreed! So the suggestion to use something only in the flat or only in the trend is only for history.

And starting from the current moment in the future there is no trend or flat, but uncertainty and statistical variation.

 
Cat Libre Black:

Agreed! So the suggestion to use something only in a flat or only in a trend is only for history.

And starting from the current moment in the future there is no trend and no flat, but uncertainty and statistical variation.

Some years ago I did a study - at what point in a trend you should enter a trade. Let's say a trend has started. How much should the price pass to enter the trade with high probability of winning? It turns out that all points are the same. No point has a statistical advantage.

 
Yuriy Asaulenko:

Some years ago, I did some research on what point in a trend to enter a trade. Let's say a trend has started. How far does the price have to go before you can enter the trade with a high probability of winning? It turns out that all points are the same. No point has an advantage.

W-o-o-o-t!

Trend length is not a statistical variable!

There is no probability distribution for it.

 
Cat Libre Black:

Pardon, it was suggested by secret.

But nevertheless the question remains open.

How to identify "only a flat".

There are many ways to identify a flat.

You can study the basic principles of recognition and create a generalised version. It will take much time, but it will be the best variant.

 
secret:
The mashka is a non-parametric indicator, it does not make any assumptions about the "function" of the price movement, but simply works (as best it can) with what is there. The trajectory has changed - the mashka has followed it.

it just builds a midline and shows the last point of it.


It's not:
Mashka is a non-parametric indicator, it makes no assumptions about the "function" of price movement, it just works (as best it can) with what is there. The trajectory changes - the waving went with it.

It makes an assumption about the function of price movement. for it, this function is a horizontal line.

It's a horizontal line:
Yes, mashka is bad at trending, her price model doesn't have the concept of a trend.

well, i'm telling you, her model is a horizontal straight line.

secret:
You just have to filter the market - take only the flat sections

You don't know at what point a flat will turn into a trend.

 
Alexander, I hope you're reading this, I found a read for you,

http://economy-ru.info/info/115012/,
In the middle, for some reason it doesn't fit, it says that the hypergeometric distribution, under certain conditions, coincides with the binomial distribution,
 
Smokchi Struck:

It simply plots the midline and shows the last point of the midline.

But it builds it disgustingly.) However, for many applications this is more than enough.

Nevertheless, EMA remains the best among other "standard" MAs in absolutely all parameters. The only problem with its smoothing period is that it really does not fit anything. Therefore it is absolutely incorrect and meaningless to compare EMA with other MAs at the same T.

 
Yuriy Asaulenko:

Some years ago, I did some research on what point in a trend to enter a trade. Let's say a trend has started. How far does the price have to go before you can enter the trade with a high probability of winning? It turns out that all points are the same. No point has a statistical advantage.

If the trend has reversed and has already gone a bit - then you won't make a profit on such entries.


And another thing. the system is the opposite of the "trend reversal" system, the pair is cointegrated.