First sacred cow: "If the trend started, it will continue" - page 13

 
Urain >> :

By this criterion you will not find any in the randomly generated one.

I will improve the criterion by taking a segment of the chart. On it we build a linear regression.

If regression is increasing we are looking for 3 minima .

We draw a line on minima (1 if possible), compare for parallels with regression (2 if it converges), it means a trend.


how much should it add up to?

 
grasn >> :

>> how much should it add up to?

If multiplier a,( y=a*x+b ) in linear regression and line plotted by minima are equal.

 
About the trend.

Keep it simple - as soon as you see a "trend" (that crap that everyone thinks is a trend), that's it, it will be over in a REALLY short time. You cannot build a strategy on this. Because just because it's over does not mean that something else will happen. :)
 
SProgrammer >> :
About the trend.

Keep it simple - once you've seen a "trend" (that crap everyone thinks is a trend), that's it, it's over in a REALLY short time. You cannot build a strategy on this. Because just because it's over does not mean that something else will happen. :)

But on the other hand, if the phenomenon exists, it can be described and perhaps its occurrence in the future can be predicted. If there is no phenomenon, there is nothing to talk about.

 
Urain >> :

If the multiplier a,( y=a*x+b ) in the linear regression and the line plotted by minima are equal.

I see, strictly equal. :о)))

 
Urain писал(а) >>

But on the other hand, if a phenomenon exists, it can be described and possibly foreseen in the future. If there is no phenomenon, then there is nothing to talk about.

Have you seen pictures like this? Like an old woman and a naked woman? That's the whole phenomenon - it has the same mechanics as hallucinations.

Try to look at wallpaper or fabric patterns - depending on your mood you will see different things but mostly a face in the sense of a face. So?

 
grasn писал(а) >>

What kind of zones are these? Where did you get them from?

-Are you a Muscovite, young lady?

-Yes! Why?

It's a joke, but...

When the price makes fluctuations, its range has a certain level to a certain extent (along the upper and lower price limits), when this range is exceeded, there is a saturation, which develops into a reverse move that forms a trend, and for different periods different range size, but if the excess in several ranges occurred in one direction, a sharp rush in the opposite direction begins ... So you have a global trend for given periods, though such situations rarely occur, but for general local trends such with

Even such local pullbacks observed clearly on the M1, which show when and how much reversal occurs... Based on these findings you can make appropriate conclusions in what direction the movement will go, there are exceptions but they are very rare, using stops... Use the Oscillator, which will filter the false reversal signals...

 
laanaa0708 писал(а) >>

When the trend is on the chart, it is time to exit the market.

Overbought and oversold is a complete abstraction.

All the so-called trends, no matter on what term, are only good to be seen on history.

The price does not know where it will be in a minute, moreover in a year.

The price may be slightly dependent on some levels.

I described above the method of determination of price in 1 minute after 5 minutes, but if you want to know where it will be in a longer timeframe, you need constant monitoring, but in any case you can find general direction half way there.

 
Urain писал(а) >>

You can draw a line through any 2 points. Not through any 3 points you can draw a line. (I think that's an axiom?)

A trend is a linear regression confirmed by a parallel line

plotted on 3 minima (or maxima). <--- Options for buy and sell.

..

However until you wait for this confirmation the trend is over.

But the trend according to which the regression was done is equal up to a certain point, but when the price begins to move away from the regression this curve begins to reverse ax+b/

 
Urain >> :

If the multiplier a,( y=a*x+b ) in the linear regression and the line plotted by minima are equal.

How reliable are you in determining the trend in this way? How many percentage errors do you have?