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
In fact, my indicator can be further simplified. We simply create in a separate window Open[0] for USDCHF and connect this indicator to EURUSD and AUDUSD. After some time (on the hourly chart - a few hours) divergence will start to appear.
In fact, my indicator can be further simplified. We simply create in a separate window Open[0] for USDCHF and connect this indicator to EURUSD and AUDUSD. After some time (on the hourly chart - a few hours) divergences will start to appear.
Just looking at your creation :) Your style has not changed, you write clearly... for yourself :)
I have two pairs - GBPUSD M5 and GBPJPY M5. Then I realized - Quark, as an experienced user, has hidden the error deeper :) I checked equation of the exponential moving average - it is correct. BUT... the code assumes, that if a new bar opens on GBPJPY (where the indicator is hovering), then a new bar will open on USDCHF (where Open[] is read from).
Is it really so? That's why the error appears gradually, in several days, because the differences need time to accumulate. I think I have explained everything clearly ?
As the saying goes: "I got angry, I was wrong, I take it all back". ;о)
Really blame the "holes" in the story. It is interesting, by the way, that I have only one "hole" from your example - 25.12.2001. But 14.03.2005 all bars are present.
I am still guessing, where 8 hours of quotes have disappeared, but that`s another story.
In any case, thanks a lot for the help. :о)
Is it really so? That's why the error appears gradually, in several days, because the differences need time to accumulate. I think I have explained everything clearly ?
My style is... I don't know. I wanted to do better. What's wrong with it? Criticism accepted. Constructive :)
Here is a new variant, without MA at all. It draws iOpen(USDCHF) and iClose.
Now about the error accumulation due to different bar open time. Formally, Open[0] is the same no matter what (it is formed at hh:00). But in practice, what if a bar on our chart has already arrived (first tick, that is), and the USDCHF (indicator currency) has not yet? Um... One would think that a properly constructed code would ask the server, but if that's not done, then yes, either the value of the previous (hour ago) Open (very wrong!!!) or the last tick value will be used (which is also not good). So maybe Roche is right.
However, there will be no error accumulation with this (it's not like we're building MAs, we're just drawing open prices).
To investigate the issue, I've put a new indicator on two charts that also draws iClose.
Let me note that even in this case, there may be divergences. For example, if last tick in one of currencies is VERY delayed, and Open on the chart comes earlier than Close on indicator's currency.
To investigate this issue, I have added a third buffer to the indicator drawing the Open of the previous bar. I think everyone will agree that these data will ALWAYS be synchronized on the watch. It's hard to imagine that one currency has Open[0], and the other hasn't Open[1] yet.
If the reasoning above is correct, then it turns out that we should be very careful when using data from another currency. It would be good if the developers would write (in a helper, or wherever) some kind of recommendation.
I'll post in about 12 hours what came out of the test.
I don't get it, explain.