For those who have (are) seriously engaged in co-movement analysis of financial instruments (> 2) - page 11
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
Temporarily diving into forexfactory.com, it's interesting and without flooding. For example, here's this thread: Co-integration in forex.
P.S. There's also a link to a blog about cointegration #regression. True, the link is broken.
Unfortunately, there is virtually nothing on ForexFactory on the subject.
From here:
You can find chewed up material on regressions, correlations and much more with examples, charts and Excel files at the links there.
The above two graphs show that you don't need to count correlation over large periods. We should build BP QC, and analyse it, and from there the cointegration will come out. I.e. a simple rearrangement of the synthetic.
No, no, there is something to the point. Maybe it's all nonsense, but it's interesting to read:
c0d3: Seems that not many people even know what cointegration means. I didn't either about a year back, since then i have developed co-integration analysis in Matlab for forex or NYSE. I have gone as far as writing a script to run cointegration on the whole 9,000,000 combination of stocks on NYSE, in about 8hrs. Similar to a hedge fund? My reseach shows, hedge funds do global analysis in about 4hrs. Forex is easier since there are limited number of time series. If you are serious about developing co-integration analysis software, buy a student version of Matlab, and download this free toolbox specific for cointegration http://www.spatial-econometrics.com/. They have a complete set of econometric analysis functions, and at first it is going to be confusing, but study ADF and CADF functions, that should give you a huge start i didn't have. It took me a while to understand how to interpret the results. My mistake is: doing analysis on high frequency data, and never actually trading the research of cointegration pairs. I might go back to it a lil later, but co-integration does exist between currency pairs, and it is not necessary USDCHF and EURUSD, you will be surprised which pairs have mean-reversion. If you are interested i can sell you the code that i have developed, it does everything for analysis for entry & exit conditions, as this strategy is exactly the same as a hedge correlation strategy, except the analysis is a bit different.
From my point of view, just the whole thing is nonsense. Stretching the market on formulas, without any justification whatsoever. There is no research into tradeability. Only different variants of "pulling" are shown.
Notice that the main issue, the difference between a real market and a random market, is not addressed at all. The pernicious influence of portfolio theories and correlation-dispersion-regressionanalysis is traced everywhere. And the more pages, matrix algebra, integral signs and models (GARCH and the like) the cooler.
And yet it's really quite simple. I haven't come across a mathematical definition of cointegration. I'll try to give one...
The bottom window shows the synthetic with the weights in the window above. Between the vertical blue lines is the synthetic construction interval, behind the lines is the OOS.
By saying that we need the synthetic not to lose its properties quickly, I meant that on the OOS it would behave similarly to how it behaved on the construction interval. See:
The horizontal dotted lines below are the levels of the synthetic's OOS on the build interval. So, what we see. The further to the right (to the left - if we were trading into the past, not into the future) of the plotting interval, the stronger the synthetic properties (horizontal channel) go. We can see that on the right border of the channel, the price of the synthetic went beyond the -SCO level (we could take another level). Why not buy the synthetic at that moment? We buy it. We wait for the synthetic to "collapse" - the price will zero out (red small triangle). Exit the position. That's it, we do not use this synthetic anymore. On each bar we will have a new synthetic. The video I mentioned above shows it all in dynamics...
And now about the clear definition of cointegration, which for some reason is nowhere to be found. See the synthetic on the construction interval (zoomed in):
Its local extrema are connected by white and green lines. The extrema are connected as follows:
1. Setting the value of N - the minimum distance of an extremum from the synthetic zero (the centre of the channel).
2. All matching extrema shall be connected in such a way that neighboring extrema are on the opposite side of zero.
Co-integration of several FI is the profit that could have been obtained by trading the synthetic from these FI on the interval of its construction. Provided that the trade size is not less than 2*N.
I.e. in our case the cointegration is equal to the sum of knees of white (green - bigger N) curve.
Then there is an immediate desire to get a synthetic with the maximum cointegration. But is it right? Above you see a synthetic that has minimal variance. Such a synthetic, as shown above, takes advantage of market relationships. Would a synthetic of maximum cointegration use market interrelationships? I don't know exactly.
The mathematical problem of finding a synthetic with the maximum cointegration belongs to the class of optimization. It is solved by numerical methods.
A simple example of BIGGER profits?
Add FI charts after OOS, though - you'll see there's a BIGGER profit to be made...
are you using the idea described in the magazine http://forum.leprecontrading.com/viewforum.php?f=92&sid=d2bc9794e202988999bcda23ebf96700 "mt4 quasiarbitrage", judging by the set of indicators on the screenshot?