Pair trading and multicurrency arbitrage. The showdown. - page 231

New comment
 
Sergey Gridnev #:
Gentlemen take my word for it 😁

there are two triangles with positive swap.

maybe there are more, but there's a mistake in the programme, I'll fix it, it doesn't matter.

Here's one of them, check it out.

Yesterday we made a robot-generator of triangles in order to check the possibility of existence of a loc with positive swap.

yes, it's possible!

 
Renat Akhtyamov #:

two triangles with a positive swap

maybe there are more of them, but there is a mistake in the programme, I'll fix it, it doesn't matter

Here's one of them, check it out.

Yesterday, a robot generator of triangles was created to test the possibility of existence of a lock with a positive swap.

yes, it is possible!

What's the ratio?

 
Renat Akhtyamov #:

Alexander's strategy is not the one he wrote about.

the trick is that when the crowd takes to trading pairs, the only cross trades predictably.



Got it. I'll continue with the dough for now.....
 
Roman Shiredchenko #:
;-) I historically consider it a robot - I will post the calculation of the screenshot here. Then I evaluate it and set limit levels like.....
As a result I have 10 levels and the number of spreads in them. I run the robot e.g. from 2023 in the tester on m15 and at the end of the test it gives out like - spreads 0-200 pips 100 pcs. 200-300 pips 89 pcs..... 500-600 pips 3 pcs.

And so I estimate the levels of downloads and inputs on the value of the current spread.
.

for evaluation I have something similar in the code and at the end of the year, for example, I look at the number of spreads (I will not put it in the code) - the algorithm is clear, who needs it if


 
Roman Shiredchenko #:

For evaluation, for now something similar in the code and at the end of the year, for example, I look at the number of gaps (I will not put it in the code) - the algorithm is clear, who needs it if


writer style for 1C ;)

your chip is done approximately like this

int step=500,sum[100];

ArrayInitialize(sum,0);

for(k=0; k<100; k++)

{

   if(k*step<=s2s1 && s2s1<(k+1)*step) sum[k]+=1;

}

 
Renat Akhtyamov #:

writer's style for 1C ;)

your trick is done like this

int step=500,sum[100];

ArrayInitialize(sum,0);

for(k=0; k<100; k++)

{

   if(k*step<=s2s1 && s2s1<(k+1)*step) sum[k]+=1;

}


;-) aha. I myself wanted and know how to realise through arrays - but then for evaluation - so decided ok will be....
 
Roman Shiredchenko #:

;-) aha. I myself wanted and know how to implement through arrays - but then for evaluation - so decided ok will be....

ok

in general, if we summarise Alexander's whole idea about bablokos, we will get the following

we take a multitool indicator

its charm is that it is still dynamic and redraws.

its redrawing is favourable because there is an algorithmic stop, even if there is a loss in case of a collapsed sliding.

A loss is possible, because pairs can diverge indefinitely.

but diverging globally (for a long time), they can repeatedly converge and diverge locally (quickly), the latter brings profit

indicator redesign was discussed in bablokos.

it was a dialogue between me and Alexander on this topic, I think it was not erased.

There is of course Alexander's improvisation on the topic of strategy improvement, but in my opinion it is better to start with the simplest (two legs) and of course on demo.

The average sliding is calculated by summing up the sliding of each bar in a long time window, divided by the number of bars in the same window.

 

Hello, everyone. The idea of taking the average has been realised. Indicator with redrawing. So far such achievements on EURUSD+GBPUSD. Lot is fixed. Checking on other pairs.

Test

 
Renat Akhtyamov #:

ok

Anyway, if we summarise Alexander's whole idea of a cash cow, it goes like this.

a multitool indicator is taken

its charm is that it is still dynamic and redraws.

its redrawing is favourable because an algorithmic stop appears, even if there is a loss in case of a collapsed sliding.

A loss is possible here, as pairs can diverge indefinitely.

but diverging globally (for a long time), they can repeatedly converge and diverge locally (quickly), the latter brings profit

indicator modification was discussed in bablokos

it was a dialogue between me and Alexander on this topic, I don't think it was erased.

There is of course Alexander's improvisation on the topic of strategy improvement, but in my opinion it is better to start with the simplest (two legs) and of course on demo.

The average sliding is calculated by summing up the sliding of each bar in a long time window, divided by the number of bars in the same window.


Opps. So far I have started with 200 pips spreads - they are the most to not miss the movement without me!!!! ;-)

And further averaging up to about 2000 - 3000 pips through 200 - 300 - 500 pips of spreads.....

Exit at the slope. Or by trawl of equity profit.

So far without flips and only towards the flop.

Opps. For an example of the logic of calculating the average spread. Perhaps it will be implemented in another robot.
 
Renat Akhtyamov #:

ok

Anyway, if we summarise Alexander's whole idea of a cash cow, it goes like this.

a multitool indicator is taken

its charm is that it is still dynamic and redraws.

its redrawing is favourable because an algorithmic stop appears, even if there is a loss in case of a collapsed sliding.

A loss is possible here, as pairs can diverge indefinitely.

but diverging globally (for a long time), they can repeatedly converge and diverge locally (quickly), the latter brings profit

indicator modification was discussed in bablokos

it was a dialogue between me and Alexander on this topic, I think it was not erased.


It can be described in a little more detail here - diverging globally (for a long time), they can repeatedly converge and diverge locally (quickly), the latter brings profit....
1...224225226227228229230231232233234235236
New comment
Reason: