Study1: multi-currency analysis for scalping and beyond
USDJPY again:
There are a lot of pictures, but it's more visual.
Now, if we observe only 5 pairs with the same change and 6 with a different change. It is still assumed that the change in rates is due to USD. But the 6th pair is skewed which should be neutralised. I.e. we open on the skewed pair in the opposite direction to the skew.
The question arises why this nasty thing sometimes does not work (pictures will be below)! The idea seems to make more or less sense, but the pictures don't make it all that rosy.
Because every currency has a life of its own, and the economy on which it is based, and finance, and politics, etc. etc. It can simply become skewed because of internal events in the country. Somebody says something, or some data comes out, or the whole damn thing burns down. It's a pain in the ass to keep track of it on the news.
Shit, then it turns out you can tell TA to fuck off in these situations.
I liked the idea because of its versatility: real multicurrency analysis + only two input parameters: the maximum radius, which includes changes of 5 pairs, and the minimum radius beyond which the "renegade" goes. Yes and multi-currency arbitrage reminds me of something, only the situations are orders of magnitude bigger and tastier.
If you sift out planned news, then TA might rule in this logic. We will have to see.
Because every currency has a life of its own, the economy on which it is based, its finances, its politics, etc. etc. It can simply be skewed by internal events in its country. Somebody says something, or some data comes out, or the whole damn thing burns down. It's a pain in the ass to keep track of it on the news.
Doubtful... The rate works out instantly on all related pairs. Has anyone seen a real skew in forex?
Doubtful... The course works out instantly on all related pairs. Has anyone seen a real skew in forex?
What connected couples? The crosses? So about them I wrote above that they can be sent far and wide as they are all known from the majors.
And the misalignments can be seen in the pictures. Only they are skews if the hypothesis of short-term multiple correlations that I described in the first post is true more often than not.
Of course, TA can be fucked if one assumes that the majors are random straying. And correlations are only a short-term visibility of something that isn't there.
In that sense, the point made here https://www.mql5.com/ru/forum/124324 has relevance.
I appeal to those who have been digging into the subject of multi-currency analysis. It is multicurrency analysis, not multi-instrument trading.
What types of multicurrency analysis do you know? The ones that I know - cluster analysis, arbitrage (one quotation source), using short-term correlations (described above).
What investigations have you conducted? MT5 has a multi-instrumental tester, but it only gives an advantage (vs MT4) in testing multi-instrumental EAs. It cannot help us to investigate multicurrency relations because we can count analysis methods on the fingers of one hand.
Anyway, let's speak constructively about multicurrency analysis.
At one of the Championships, Rosh shared his vision of the characteristics an ideal strategy should meet. One of the characteristics is multicurrency analysis for trade management and MM...
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
You agree to website policy and terms of use
Let's take 6 major symbols (EURUSD, GBPUSD, AUDUSD, USDJPY, USDCHF, USDCAD) and look at relative changes of their rates.
If, for example, during the last 10 minutes, all 6 major rates have changed by 0.05% (in one direction relative to USD), then we assume that USD has played the main role.
Now, if we observe only 5 pairs with the same change, and 6 with a different one. We still assume that the change of rates was due to USD. But the 6-th pair has a skew which should be neutralized. In other words, we open on the skewed pair in the opposite direction to the skew.
This is just an idea-hypothesis of "pricing".
I suggest looking at some pictures of such situations, which are now automatically found and saved.
(AUDUSD is skewed here).
The question arises, why does this nastiness sometimes not work (pictures to come below)! The idea seems to make more or less sense, but the pictures don't make things so rosy.
I will say right away that the crosses are not fucking necessary, as they are all easily obtained from the majors. That's why the assumptions like JPY is sold through EUR and not USD, that's why you can't see it in the majors are groundless. Because the USDJPY and EURUSD majors fully reflect the situation on EURJPY.
Yes, why I wrote scalping. Because the idea seems reasonable only on small timeframes. Then the probability of influence on rates of only one currency is much higher. But it is not a fact.
I want to hear constructive thoughts in a sea of flooding...