Pair trading and multicurrency arbitrage. The showdown. - page 67
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
A cent is a very professional cent, pro cent :-)
The humour joke is good, funny.
Understanding the method of calculating your pro cents would help you understand the nature of the significant correlation between your graphs.
The humour joke is good, funny.
Understanding the method of calculating your pro-rates would help you understand the nature of the significant correlation between your graphs.
see page 49 of https://www.mql5.com/ru/forum/448777/page49#comment_49894424.
and somewhere else there was a graph about the relationship between currencies, that in harsh reality it has all the nodes on the same line.
see page 49 https://www.mql5.com/ru/forum/448777/page49#comment_49894424
and there was somewhere else about that graph of relations between currencies, that in harsh reality it has all nodes on one line
Economists usually use the consumer basket as a unit. In our field we can still talk about the basket of currencies.
Actually, the original question arose from the fact that if dollars are invested in dollars, there will always be a return in dollars of exactly zero per cent - a horizontal straight line, which you do not observe).
Economists usually use the consumer basket as a unit. In our field we can also talk about the basket of currencies.
Actually, the original question arose from the fact that if dollars are invested in dollars, the return in dollars will always be exactly zero per cent - a horizontal straight line, which you do not observe).
You're not looking carefully.
on the picture https://www.mql5.com/ru/forum/448777/page66#comment_50020066 top part...the horizontal line 0 is exactly the return on investing dollars in dollars.
you're not looking carefully
in the picture https://www.mql5.com/ru/forum/448777/page66#comment_50020066 the top part...the horizontal line 0 is just the return on investing dollars into dollars.
Ah, we invest dollars at the top and gold at the bottom. Then the perfect negative correlation between the price of gold in dollars and dollars in gold is clear) For the rest not so perfect.
To check the advance, it is easiest to calculate the extremum of correlation of two series depending on the shift between them (somewhere on the forum I calculated for silver and gold). You can also look at the statistics of precedence on the tops of zigzags. As a rule, nothing interesting.
Ah, we invest dollars at the top and gold at the bottom. Then the perfect negative correlation between the price of gold in dollars and dollars in gold is clear) For the rest not so perfect.
To check the advance, it is easiest to calculate the extremum of correlation of two series depending on the shift between them (somewhere on the forum I calculated for silver and gold). You can also look at the statistics of precedence on the tops of zigzags. As a rule, nothing interesting.
how you like to calculate correlations on anything that comes to hand :-))
To calculate the correlations of "the easiest thing" - you will have to remove trends again/ again, take into account trading sessions, daily volatility, schedule shifts (clock translation in different places). This is only a first approximation.
But the answer is known in advance: gold prices (any asset) started to move, shifts/shifts between others (fluctuations of individual pairs) started. It's the same thing, one-step at a time.
It is just that on a single pair it is less visible and sometimes later becomes noticeable and obvious
By the way, most likely, almost certainly:
(only about currencies!) the collapse of correlations of pairs (there was a positive high one, and then suddenly there was a small or inverse one) is closely related to the limits on profitability. All pairs move at the same pace and correct when it is exceeded.....
how much you like to calculate correlations on anything that comes to hand :-)
In order to calculate correlations of "what is easiest" - you will have to remove trends again/ again, take into account trading sessions, daily volatility, schedule shifts (translation of clocks in different places). This is only a first approximation.
But the answer is known in advance: gold prices (any asset) started to move, shifts/shifts between others (fluctuations of individual pairs) started. This is one and the same thing, one-stage.
It is just that on a single pair it is less visible and sometimes later becomes noticeable and obvious
.
Matstat is full of examples when intuition lies, so it is always better to calculate a little than to assume a lot.
Simple models are enough to start with, and if they show something interesting, you can try to refine and complicate them somehow.
.
Matstat is full of examples when intuition lies, so it is always better to calculate a little than to assume a lot.
matstat (and the local MoD) is full of examples where they apply methods inappropriately to anything.
abstract models on arbitrary data, without the slightest logic and revealed interrelationships...In the hope that the methods themselves will somehow find everything, reveal everything, and the philosopher's stone will be obtained
Alchemical recipe: "you just need to put arbitrary 2-3-N rows in the right pot and stir according to the instructions" ; the result is predictable, but you can change the pot and stir at a different pace.
all the MO present on the site, it's about changing pots and stirring instructions. Doesn't even try to really look for connections, build theories, even just look at the nature of the series. Doesn't rely on anything.
Pure uncontaminated belief "there are MOs and neural networks, they will do anything to stop them".
matstat (and the local MoD) is full of examples where they apply methods unmercifully to anything.
abstract models on arbitrary data, without the slightest logic and revealed interrelationships...In the hope that the methods will somehow find everything themselves, reveal everything, and the philosopher's stone will be obtained
Alchemical recipe: "you just need to put arbitrary 2-3-N rows in the right pot and stir according to the instructions" ; the result is predictable, but you can change the pot and stir at a different pace.
all the MO present on the site, it's about changing pots and stirring instructions. Doesn't even try to really look for connections, build theories, even just look at the nature of the series. Doesn't rely on anything.
Pure uncontaminated faith "there are MOs and neural networks, they will do anything to stop them".
That's a good thing, less competition)