Pair trading and multicurrency arbitrage. The showdown. - page 302

 
Vitaly, on average, how many profitable trades do you have for every 10 trades you make?
 
Sergey2134 #:
Vitaly, on average, how many profitable trades do you have for every 10 trades you make?
10
 
Vitaly Muzichenko #:
The calculation of ++++ and ---- is based on correlation; if the pairs diverge at any point, the calculation may be incorrect – this applies if + and -, or ++ and -- – match. Please double-check visually by reversing the second pair.

Here’s an example where the reversal produced a false signal


After the reversal, the divergence percentage will change; if it is less than 60%, ignore it



You need to remove the reversal.

It seems logical that these are allied pairs, and one should be sold whilst the other should be bought.

It’s now possible to get a lot of information from AI: ask a question about specific pairs and get an answer as to whether they’re allies or mirror pairs.

If they’re mirror pairs, then two buys or two sells.

--

The system is simple, but it requires attention and a bit of logical thinking.

 
Vitaly Muzichenko #:

The system is simple, but it requires attention and a bit of logical thinking.

Yes, it’s simple. But there are some nuances. Today I opened a CADJPY+EURNZD position on the demo account. Well, you can see from the measurement that the spread is 200 pips. So why take 10 pips out of 200 pips? Maybe there’ll be more pips of movement after all? )

Files:
03_08_2026.png  30 kb
 
Sergey2134 #:

It’s simple, really. But there are a few nuances. Today I opened a CADJPY+EURNZD position on a demo account. Well, you can see from the measurement that the spread is 200 pips. So why take 10 pips out of 200 pips? Maybe there’ll be more pips of movement after all? )

It’s not 200 pips; the scale is fixed from -1 to 1 to standardise the ratios across all instruments.

From the chart and the lot sizes, you can see that no quotes are being loaded for the EURNZD pair = it’s almost a straight line. Switch the timeframe forwards or backwards

Or the second option: a shorter timeframe. There was significant movement in the yen, so the calculation is currently incorrect.


 
Oh, thanks. I’ve updated it. But a fixed scale like that is a bit confusing if you’re measuring the points by hand. I was actually planning to measure the spread manually.
 
Sergey2134 #:
Oh, thanks. I’ve updated it. But a fixed scale like this is a bit confusing if you’re measuring the points manually. I was actually hoping to measure the spread by hand.
Is there another way to link symbols with different prices to a single scale?
 
There’s no other way to set it up. Perhaps, when taking a position on Teika, it would be better to consider the percentage of the spread narrowing? Let’s say we opened a position on a signal and the spread has narrowed by 50 per cent. In that case, ‘Close All’.
 
Sergey2134 #:
There’s no other way to set it up. Perhaps, when taking a position on Teika, it’s better to look at the percentage of the spread narrowing? Let’s say you opened a position on a signal and the spread has narrowed by 50 per cent. Then ‘Close All’.
Have a look at this – I haven’t managed to close for a profit any other way.
 
Vitaly Muzichenko #:

Here’s an example of a false alarm


After the reversal, the deviation percentage will also change; if it is less than 60%, ignore it



The reversal needs to be removed.

It makes sense, in a way, that these are allied pairs, and you need to sell one and buy the other.

It’s now possible to get a lot of information from the AI; you can ask a question about specific pairs and get an answer: allies or mirror pairs.

If they’re mirror pairs, then that means two buys or two sells.

--

The system is simple, but it requires attention and a bit of logical thinking.

I’ll add my two cents: EURAUD 0.1 sell, GBPNZD 0.8 buy. Based on the current position, profit is $50, loss is -$150 – not the best time to enter a trade. EURAUD 0.13, GBPNZD 0.1 – these show more stable performance. That’s according to my maths; I’m not claiming to be an expert.