The best trend-finding indicator

 
In your opinion what is the best trend determining indicator among known indicators, i.e. which one captures the trend range as much as possible?
In my opinion it is an average based on JJMASeries algorithm ('Efficient algorithms of averaging with minimal lag and their use in indicators') with long settings.

What do you use?
 
liveexpert:
In your opinion what is the best trend determining indicator among known indicators, i.e. which one captures the trend range as much as possible?
In my opinion it is an average based on JJMASeries algorithm ('Efficient averaging algorithms with minimal lag and their use in indicators') with long settings.

What do you use?

Difficult. Especially when highs and lows are increasing.
In my opinion, impossible. Either only on the basis of eliot waves.

By the way, is there any software that is able to detect waves and their subdivisions at the market?
It is difficult to write one yourself, but it would be easier to modify a ready-made one.
 
liveexpert:
In your opinion what is the best trend determining indicator among known indicators, i.e. which one captures the trend range as much as possible?
In my opinion it is an average based on JJMASeries algorithm ('Efficient algorithms of averaging with minimal lag and their use in indicators') with long settings.

What do you use?
1) The answer will at least depend on the definition - "What is a trend" :)
There is no single correct definition - accordingly there cannot be an optimal trend detection algorithm.

2. Why do we need an average without a lag to determine the trend?
After all, the trend is the direction of price change over some time.
So, a time lag is a MUST for trend detection .
It is impossible to identify any trend without it - this is a matter of principle.

3. "Detecting a trend" and "Predicting the presence of a trend" are fundamentally different things.
You probably mean the second one, and it is either impossible, or if it is possible, it will be different for different securities, and not constant in time. For example, on some stock a Gap up may be a sign of the start of growth with XX% probability, on another stock it may be a sign of decline. And this behaviour may change depending on external factors. "Determination" is a measurement by past data, while "forecasting" is a prediction for the future. These things are often confused for some reason.
 
Nothing like Heiken Ashi
 
Mak:
liveexpert:
In your opinion, which of the known indicators is the best one that detects the trend, i.e. captures the range of the trend as much as possible?
In my opinion, it is an average based on JJMASeries algorithm ('Efficient algorithms of averaging with minimal lag and their use in indicators') with long settings.

What do you use?


1) The answer will at least depend on the definition - "What is a trend" :)
There is no single correct definition - accordingly there cannot be an optimal trend detection algorithm.

2. Why do we need an average without a lag to determine the trend?
After all, the trend is the direction of price change over some time.
So, a time lag is a MUST for trend detection .
It is impossible to identify any trend without it - this is a matter of principle.

3. "Detecting a trend" and "Predicting the presence of a trend" are fundamentally different things.
You probably mean the second one, and it is either impossible, or if it is possible, it will be different for different securities, and not constant in time. For example, on some stock a Gap up may be a sign of the start of growth with XX% probability, on another stock it may be a sign of decline. And this behaviour may change depending on external factors. "Determination" is a measurement by past data, while "forecasting" is a prediction for the future. These things are often confused for some reason.



I'm talking exactly about determination, I'm not thinking about prediction =) the delay is clear, the question is which indicator shows the least delay.
 
liveexpert писал (а):

I'm not thinking exactly about the definition, the forecast =) the delay is understandable, the question is which indicator shows the least delay.



I think what Mak is trying to say is that if the lag is minimal, the "trend indicator" will often change its readings, and that is not good :)
 
Mak, sashken and which indicators (or combination of indicators) do you use to determine the current trend?
 
Delay is necessary to estimate the trend by definition.
Trend estimation is actually estimating the sign of the price derivative over time.
The most natural estimate is the smoothed derivative - for example the Trix indicator.
There are many other variants, but all of them actually estimate the sign of the derivative.
I liked the Asctrend indicator.
But 2 muwings is just as good.

A good indicator is MACD, it evaluates not only the first but also the second derivative.
The second derivative is a leading indicator of a trend change.
In other words, it allows to notice a trend change trend,
while the old trend is still there.

Generally speaking, there is no universal solution or the best indicator.
You have to choose it yourself according to your specific task.
 
Here is another methodology that I find curious:

Combining PSAR and ADX

http://www.kroufr.ru/content/view/1972/124/
 

In fact, I've only seen one real expert indicator in person, which detects the presence of a trend here. It's the first one on the page after registration.

 
Do you mean _xMeter_mini_v.1.3?