Some signs of the right TCs - page 12
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 was asking that one does not see the practicality for constructing a TS, because for different tools one still has to optimise and find different parameters, in practice different tools are unique, there are no tools obtained from others by similar transformations, except synthetics.
Real symbol or synthetic is a convention. Well they will translate GBPEUR to you in the Terminal. For the TS it should not differ from EURGBP. The main thing is to broadcast a relative change that is traded.
If the TS trades a relative change, it automatically falls under the discussed invariant.
To optimize EURGBP and GPUEUR means to do a double work. Optimizing EURUSD and EURGBP is different work. It is therefore logical to do it as with the other synthetics.
It seems that there is a search for a "grail", a universal market formula and an imposition of misconceptions (or a pre-publicity of unreleased products a la "car tester at a converted price range").
I don't create threads very often. They allow you to draw some conclusions for yourself as to what opinions there are on the subject. How many points of view coincide. It's good if some interesting things are voiced.
For example, the time inversion is excellent, as it is brought to practical theoretical research. Well, theory helps to roughly choose the direction of one's efforts for greater impact in practical terms.
reducing the likelihood of fitting, that's all.
The "rightness" does not guarantee profitability, and vice versa, but it is easier to work with the "right" TS.
This is what I really don't understand. A simple example: a TS with a hard stop. 1. if the principle is not respected, the stop is set in points 2. if the principle is respected, the stop is calculated so that the result does not change for the specified transformations. Okay, we optimize them using whatever methods and obtain an optimal stop = 400 points in the first case, or 0.4% in the second, for the period we are interested in.
Ok, then we multiplied by two and check - in the 2nd case, the system shows the same result without changes, in the 1st case it is of course not, we have to re-optimize and get the stop at 800 points.
Further, the question is: WHAT DOES THIS MEAN? :)
In essence, the TS remains the same, its capabilities are the same. If it is subject to manipulation, it will be subject to manipulation, if we do not take measures to stop it. Measures have nothing to do with the correctness of the TC in that sense.
s.s. As I understood from the discussion - no one has any idea what it gives, it's just that fxsaber made a generalizing conclusion, it's correct, someone understood it, someone didn't, but in practice it means little.
Reducing the likelihood of fitting, that's all.
I haven't voiced this on purpose, as it's easy to cause flooding on dozens of pages.
The "rightness" does not guarantee profitability, and vice versa, but it is easier to work with the "right" TS.
Sometimes it's not just easier, but necessary if you want to do large-scale research.
Whether it is a real symbol or a synthetic one is a convention. So you will have GBPEUR in your terminal. For the TS it should not differ from EURGBP. The main thing is to broadcast the relative change, which is traded.
If the TS trades a relative change, it automatically falls under the discussed invariant.
To optimize EURGBP and GPUEUR means to do a double work. Optimizing EURUSD and EURGBP is different work. Therefore it is logical to do it as with other synthetics.
I agree with that. Apparently nothing else does.
ap: Maybe I just don't see any other differences except for expressing all system parameters by BP function, and not by "absolute" steps - points.
And by the way even then - it is possible that there are patterns that depend on these very "absolute changes". I.e. dependence exists and can theoretically be found and traded, but systems "correct" from this point of view cannot do it. So there are two sides to this correctness.
I agree with that. Apparently nothing else does.
ap: Maybe I just don't see any difference other than expressing all the system parameters with a BP function rather than "absolute" steps - points.
Imagine you were given an impersonal price series and nothing else. There should be no obstacles to optimizing the TS on such a series.
And by the way, even then - it is not excluded that there are patterns that depend on these very "absolute changes". I.e. the dependence exists and it can be theoretically found and traded, but the systems "correct" from this point of view cannot do it. Thus, this correctness has two sides as well.
I often hear about breakthrough of a psychologically important price level (preferably, of course, "round". It should be a multiple of the decimal notation. After all, the market cannot depend on the number of homo sapiens' fingers) and other clevernesses. For example, EURUSD breaks down 1.25 (or 1.33), and USDEUR breaks up 0.8 (or 0.75), respectively. It is clear that psychologically the second action is much stronger than the first one, because the level is "rounder".
And so everywhere...
fxsaber:
After all, the market cannot depend on the number of fingers of a homo sapien) and other cleverness.
If this is sarcasm, it is for nothing, it depends in fact. not only that the principle stated in the first post does not work for the TC exploiting it (another one does)
I admit that the more popular the symbol, the more likely the influence of horoscope buffs, including decision makers at the Central Bank.
Yes, in this scenario highly intelligent phenomena are likely.
Extremes have always been evil. Round levels are round levels. De-personalise the row, hide those circular levels and you won't find the dependencies that were on them anymore, right? That's evil.
But you can weed out some of the "wrong" TCs?
It seems to me that only very "dumb" TSs, not written in a flexible way for specific conditions, will be eliminated.
The rest of the "normal" TS will just need to be reoptimized.
Once again, extremes are evil. You need a golden mean. You shouldn't throw the baby out with the water.
The correctness given is needed, as you say, for automation of large-scale studies, and I argue that it is more important for such studies to define Boundary Conditions (BC).
Once again, extremes are evil. There has to be a middle ground. You shouldn't throw the baby out with the bathwater.