For those who have (are) seriously engaged in co-movement analysis of financial instruments (> 2) - page 3
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
I would still like to add - all the same - in the end it will all come down to a special case of Pair trading :) -
In order to get a little fixed equity - you would have to do - 1 currency against a basket - or a basket against a basket... and so... recalculating the whole group of currencies every bar is not practical...
Thanks for the link!
There the author gives the method via a one-sided linear multivariate regression.
The author also shows the results in a video.
I am posting similarly one of my results.
What is the difference between the calculation of weights ( number of lots?) of the synthetic tool in your IND_Recycle and the author's tools in the links above?
Why are you only looking at currencies? In my opinion, the most interesting approach would be to create a synthetic based on negatively correlated instruments like dollar index - gold CRB-Index - bonds, etc. I think, in general, one should be guided by inflation-deflationary processes and only on a large time scale.
In general there is no need to create a synthetic with rigidly defined channel type indicators, etc. It is very difficult and probably not even possible. Such a synthetic would not be stable over time. But it is quite capable of generating a global flat. What prevents you from running any flat TS on such synthetics?
What is the difference between the calculation of weights ( number of lots? ) of the synthetic instrument in your IND_Recycle and in the author's tools on the above links?
By way of example:
The author of the paper under discussion has the current symbol (on which his code runs) expressed through the others by applying a multivariate linear regression. It was shown above that linear regression is one-sided. This means that if you run its code on different characters, you will get disproportionate weighting factors.
Recycle uses a completely different approach, so there is no ambiguity.
Why are you only looking at currencies? In my opinion, the most interesting approach would be to create a synthetic based on negatively correlated instruments like dollar index - gold CRB-Index - bonds, etc. I think, in general, one should be guided by inflation-deflationary processes and only on a large time scale.
Currencies are given in the description. In fact you can consider FOREX+NYSE+NASDAQ+BP Equity strategies.
In general, you don't need to create a synthetic with rigidly defined channel type indicators, etc. It is very difficult and probably not even possible. Such a synthetic would not be stable over time. But it is quite capable of generating a global flat. What prevents you from running any flat TS on such synthetics?
By the way, here is an idea about BP Equity N strategies: it is no secret that it is very difficult to create a strategy that will steadily raise the deposit. But it is much easier to invent a stable strategy, which does not give a profit, but does not lose money. If there are N such strategies, we will form a very narrow Equity bell around zero. Following this logic, the equity of any strategy should unfold within it, on which we can bet.
And where is the guarantee that the synthetics and weights created are not a fit for the graph we need so much? To be honest, I don't really understand the principle behind the weighting factors. Logically they should maintain the volatility of portfolio elements at the same level. But in that case we need dynamic coefficients, because volatility changes over time.
Alexander, how did you manage to combine several charts in one window? For some reason I was not able to do so.
And where is the guarantee that the synthetics and weights created will not be a tweak to the timetable we so desperately need?
There is no guarantee. If these equity are random walks, i.e. independent at all, then it will be the purest fit.
Frankly, I don't really understand what principle the weights are chosen on. Logically they should maintain the volatility of the portfolio elements at the same level. But in this case we need dynamic coefficients because volatility changes over time.
They are dynamic coefficients.
Alexander, how did you manage to combine several charts in one window? For some reason I was not able to do so.
Facts.
There are no facts. But you can imagine the sum of all the pairs. It has the shape of a horizontal line. Now remove something from the sum...
The facts are there. We need the synthetic instrument to make some kind of economic sense (the sum of all pairs is what? and if I add up the prices of 40000 instruments?)
And the facts are in the most prominent places. Instead of thinking it's better to take it on the chin, eh, sanyooooook?
The facts are there. You need a synthetic instrument to make some kind of economic sense (the sum of all pairs is what? What if I add up the prices of 40000 instruments?