An obvious way of predicting quotes for colleagues - page 2
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
Obviously, September 3 is a day off for the USD, on Friday 31 these pairs synchronously made lows at the end of the day, and it is somehow not farsighted to predict/expect anything without a major baboon in the economy. And if you look at the value, weighted dollar index, it is near the value of the high of Friday 31, i.e. the previous trading day. Another 500-page thread with projects to bend xforex.
We are trading the euro and pound difference in its pure form, the fact that it is represented through the dollar as a quote currency in my pictures is not the point. It would be the same with any other quoted currency, e.g. EURJPY and GBPJPY.
The dollar has nothing to do with it at all.
We are trading the euro and pound difference in its pure form, the fact that it is represented through the dollar as a quote currency in my pictures is not the point. It would be the same with any other quoted currency, e.g. EURJPY and GBPJPY.
The dollar has nothing to do with it at all.
You wrote above that synthetic does not tradehttps://www.mql5.com/ru/forum/277305#comment_8556518 , please, on the example of two days(August 30,31) indicate the points for trading with your system. (How can it not trade if the dollar is used to settle and the warehouse of dollars and other currencies in the usa is almost not involved in trading?)
You wrote above that synthetic does not tradehttps://www.mql5.com/ru/forum/277305#comment_8556518 , please, on the example of two days (30,31 August), indicate the points for trading on your system. (How can it not trade if the calculation is through the dollar, and the warehouse of dollars and other currencies in the usa is almost not involved in trading?).
But the conclusion what to sell and what to buy cannot change, right? Now we sell euro (a pair to dollar, if you like) and buy pound (a pair to dollar, if you like), so I will consider my profit in dollar pips. You do realize that if I trade the euro to pound ratio - at once - without the third currency involved - then the result will be in pound pips, and it relates to dollar pips as it is known at the moment: approximately as 1.3.
It is all the same, even in granny's cakes, what time to sell/buy at 13:00 or 16:45, etc.? The trading solution is a trigger and it triggers instantly in our theoretical case, this moment has a date-time and the question according to date-time, what is buying and selling at a certain moment is also a question?
Dear colleagues, Good afternoon.
I'm sure everyone has repeatedly noticed that some instruments are correlated.
For example, EURUSD and GBPUSD.
Aq and Bq curves correlate on the specified interval of 576 bars with Pearson's linear correlation coefficient over 0.9994.
What you suggest is pair trading, a special case of statistical arbitrage. And the correlation is not needed here. Moreover, even the absolute correlation with kk=1.0 does not imply the possibility of its trading, it just means that the instruments move in the same direction. At the same time the scales of their movements can significantly differ and the instruments will diverge further and further.
For pair trading we need cointegration, i.e. variance reversion. And it is much more difficult to arrange. By the way, cointegrated BPs may have very low or no correlation at all.
Trading based on the correlation will give profit in some area, like when using averaging or Martingale, but there is no statistical advantage in it.
An example of how BPs can infinitely diverge in case of high correlation.
An example of cointegrated BPs with low correlation. These are exactly the ones of interest for trading.
Colleagues, the time is 9:15 am on September 6. Let's save EURUSD5 and GBPUSD5 charts and process them.
Namely, we will examine the evolution of EURUSD and GBPUSD charts since my previous post, for the sake of clarity we will show everything on the time interval of 5 days (5*288 bars M5).
What do we see? The stated profit of 70 pips has been easily taken. More clearly in terms of difference (and ratio).
We can see how the difference dropped from -0.1250 to -0.1350. Note that this happened in an abrupt spurt. This is often the case: tensions (mismatches) accumulate slowly, relax - abruptly.
We can also see that at the moment we should: sell EURUSD, at the same time buying GBPUSD. Expected profit = 48 pips.Let's show it close-up, on a 2 day interval:
... absolute correlation with kc=1.0 does not at all imply its trading possibilities, it just means ... instruments will diverge further and further over time ...
You are quick to grasp the obvious. That is why I say that it is the correlation of less than 1 that is important, close but less - it allows the additional curves not to diverge further and further over time as you fear...
You are quick to grasp the obvious.
I'm not a fast learner, but I've been studying this topic for 10-15 years, both theoretically and practically. Fortunately now MT5 presents the possibility of multi-currency testing, you can save yourself a lot of time and possibly money by testing your idea on history.
P.S. Regarding the divergence/divergence of instruments. Yes, in 70-80% or even 90% of cases depending on the aggressiveness after the divergence, they will converge, but the remaining 10-20-30% of large losing trades will eat all the accumulated profit and pull the account into deficit. Not immediately of course, but the black swan will do its job.
... the important thing is precisely that the correlation is less than 1, close, but less - this allows the additional curves not to diverge further and further with time as you fear...
This is a fundamentally wrong concept. You seem to have the wrong idea about correlation. It does not imply the reversibility of BP differences, but the co-direction of motion vectors, and the modules may differ significantly, and the cumulative difference may increase.
Here is an example at QC=0.98
I'm not a fast learner, but I've been studying this topic for 10-15 years, both theoretically and practically. Fortunately now MT5 presents the possibility of multi-currency testing, you can save yourself a lot of time and possibly money by testing your idea on history.
P.S. Regarding the divergence/divergence of instruments. Yes, in 70-80% or even 90% of cases depending on the aggressiveness after the divergence, they will converge, but the remaining 10-20-30% of large losing trades will eat all the accumulated profit and pull the account into deficit. Not immediately of course, but the black swan will do its job.
If I tell you that it is impossible, because it is stated in the algorithm for constructing additional curves Aq and Bq, due to which Aq and Bq have some additional symmetries (for example, Aq+Bq = A+B), would you be satisfied?