Why is the normal distribution not normal? - page 16

 

EURCHF is a highly liquid instrument, but potentially low yielding, in other words a weak signal/noise

because continental European currencies are strongly correlated : EUR, CHF, SEK, and DKK is tightly bound to EUR

it is called falsification of research or at least not objectivity

Neutron you have a wasted talent not paid for ... the Rothschild team appreciates such specialists and pays well for global warming theories ))))

 
bank >> :

EURCHF is a highly liquid instrument, but potentially low yielding, in other words a weak signal/noise

because continental European currencies are strongly correlated: EUR, CHF, SEK, and DKK is tightly bound to EUR

it is called falsification of research or at least not objectivity

Neutron you have a talent for wasting ... the Rothschild team appreciates such specialists and pays well for the global warming theory ))))

I don't think Sergei falsified anything. It seemed to be about the correlation of increments within the same time series (any given quote). He just showed the result, which is obtained by clear methods, examining different scales

 
grasn >> :

I do not think that Sergei falsified anything.


I marked it as the maximum and the minimum is not objectivity.

>> there's a joke in every joke.

 
bank писал(а) >> Highly liquid instrument, but potentially low-yielding, in other words weak signal/noise

because continental European currencies are highly correlated: the euro, the franc, the Scandinavian kroner and the DKK is tightly tied to the euro

it is called falsification of research or at least not objectivity

Neutron you are wasting talent not paid for ... the Rothschild team appreciates such specialists and are well paid for global warming theories ))))

Do you think this is a chart of a liquid instrument?

And you can see the trading volumes of this cross, where they are given (on Oanda for example). They are not comparable to the volumes of major currencies or Yen crosses. And the spreads on them are big at night on the market. There may be different reasons for the instrument being illiquid, including those mentioned above.

 
grasn >> :

I do not think that Sergey has falsified anything. It seems to have been about the correlation of increments within one time series (any given quote). He simply showed the result, which is obtained by clear methods, investigating different scales

Thank you, Sergey, for your support.

bank wrote >>
>> because continental European currencies are highly correlated: the euro, the franc, the Scandinavian kroner and the DKK is tightly pegged to the euro

There is a mathematically correct way to reconstruct currency indices from the analysis of major instruments. The main thing is to have more input data.

Recently, especially for this purpose I've been collecting data on all currency pairs in Alpari during 1 or 2 months. Here is the dynamics for the indices:

You can see that the world currencies do correlate with each other, some anticorrelate. But it is difficult to understand anything clear from this mishmash of lines. It is easier to find the pair correlation coefficients between all the indices and look at them.

From the table we can see that only Swiss is "intentionally" attached to the Jew. The GBP is weaker, because part of the time it "chases" the JPY. The euras and the bx do not see eye to eye.

Interestingly, the existence of a correlation between two assets "chasing" one leader does not yet indicate a real connection between them (it turns out to be indirect). It's like in a team: some people who have no interest in each other but nevertheless have to work in sync, obeying the general plan of the company. So there are meaningful (or true) dependencies and imaginary (mediated) ones... How can they be separated?

 
Neutron писал(а) >>

Thank you, Sergey, for your support.

There is a mathematically correct way to reconstruct currency indices from the analysis of major instruments. The main thing is to have more input data.

Recently, especially for this purpose I've been collecting data on all currency pairs in Alpari during 1 or 2 months. Here is the dynamics for the indices:

You can see that the world currencies do correlate with each other, some anticorrelate. But it is difficult to understand anything clear from this mishmash of lines. It is easier to find the pair correlation coefficients between all the indices and look at them.

From the table we can see that only Swiss is "intentionally" attached to the Jew. The GBP is weaker, because part of the time it "chases" the JPY. The Yevra and the Bax do not see eye to eye.

Interestingly, the correlation between two assets "chasing" the same leader does not mean there is an actual connection between them (it is mediated). It's like in a team, some people who have no interest in each other but nevertheless have to work in sync, obeying the general plan of the company. So there are meaningful (or true) dependencies and imaginary (mediated) ones... How can they be separated?

Have you tried the same but with a shift in time, e.g. a day?

 
Avals >> :

Do you think this is a chart of a liquid instrument?

And the trading volumes of this cross can be seen where they are given (on Oanda for example). They are not comparable to the volumes of major currencies or Yen crosses. And the spreads on them are big at night on the market. There may be different reasons for the instrument being illiquid, including those mentioned above.


>> liquidity is determined by the relative spread.

Spread (average for market makers) is less with francs than with pound or yen

 
Neutron >> :


There is a mathematically correct way to reconstruct currency indices from analysis of the underlying instruments. The main thing is to have more input data.



enough dollar pairs instead of indices
 

paukas писал(а) >> А вы не пробовали тож самое проделать но со сдвигом во времени, например на сутки?

Of course I did, and not just for a day, but for any time interval... And I also plotted it out.

The result is as follows:

Statistically significant correlations are retained only when there is no shift.

When shifting by ONE tick, the amplitude of correlations decreases approximately 2-5 times, and after shifts of more than ONE tick - there's nothing to catch!

It seems that arbitrage based on the ineffectiveness of the market is possible only within a tick, and is not available to us in principle. We need completely different platforms and not VCs capable of placing orders in the market within milliseconds.

 
bank >> :


liquidity is determined by the relative spread


It's more of a liquidity-dependent spread