Once again, arbitrage, pair trading. - page 16

 
Darirunu:

I was thinking the same thing, but I ended up thinking, a rock, a ball or just paper ... What difference does it make? If you throw it at the right angle it will fly to the maximum.

Well, I wrote there - the most important thing is the trend, the angle is an easy thing to determine.

Although, I wrote in the forecasts branch and gave a link to the link - how to draw the trend line and the noise level, which should be discarded, appears there as the only parameter

But no one was paying attention...

 
Renat Akhtyamov:

Well, I wrote there - the main thing is the trend, the angle is an easy thing to determine.

Although, I wrote in the forecasts branch and gave a link to draw the trend line and the noise level appears there as the only parameter

But no one was paying attention to it...


I reject the noise by the angle, not by plotting the angle.

 
Darirunu:

Probably wrote it again like in the book of nostadamus ))) I reject the noise by the angle itself .I do not draw the angle rejecting the noise .

You don't draw an angle, you draw a trend by discarding the noise and then define the angle

that's what I wrote.

 
Renat Akhtyamov:
You don't draw the angle, you draw the trend by discarding the noise and defining the angle

That's what I wrote.


All right, let me ask you another way... What do you need to draw an angle?

 
Darirunu:

What do you need to draw an angle?

I was building a trend straight from the pivot, that's what I wrote.

I've been doing this since about 1.2 years ago. Now it's not relevant to me.

 
Renat Akhtyamov:
The trend line was built right from the pivot. I wrote it.

I mean, one price is enough ... It's just that in your own indicator the corner is always running behind the price so to speak ... Almost like a muving...

I'm going to bed now. The pound will get to 1.3525 by itself.

In the morning there will be a new fairy tale...

 
Darirunu:

I mean one price is enough ..It's just that in your turkey the corner is always running behind the price so to speak .

a triangle, can you imagine?

It actually draws it.

I don't remember what's running there...

increase the noise parameter so that every sneeze of the price doesn't affect it...

 
Darirunu:

Pay attention to the degrees of the corners...

But you're going to have a question of when and where to build them.

That's what you really need to think about and you'll get a new look at the graph.


Degrees of angles? How and how to measure it, when time is on the X-axis and price is on the Y-axis?

And the slope of the trend changes when the vertical and horizontal scales are altered.

 
Alexander Sevastyanov:

Degrees of angles? How and how can they be measured when the X-axis is time and the Y-axis is price?

And the slope of the trend changes when you change both the vertical and horizontal scales.


You need to look at the relative ones, and not degrees, but the linear regression coefficient.

 

after some tests (10 days) i used to trade on penny to see what happens

i just opened trades on spread and went about my business, or periodically i looked

sometimes 1 spread was opened and sometimes others were added to it (i.e. portfolio was traded selectively and not the whole portfolio at once)

This type of trading is the calmest and least stressful one, where you almost never have to think. I wonder why I haven't noticed it before.

There were no force majeure in the market during this period. Not all signals were traded, you can say "out of the blue".

At the end of the day, all spreads converged, and if they didn't, they were hedged with other spreads from the portfolio, and they finally converged

towards the end began to increase a bit the lot, maybe even to the point of serious trading

I will check it further with the bot



Reason: