Combining trend and flat strategies into one TS = grail? - 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
then there is no point in combining them. Each part is a grail. And since we now have two of them, one - to choose from - can be profitable for the suckers)))
First you need to find correlations among the strategies. Then trade them through weighting coefficients for each strategy.
Just an imho. You have to get the gist of it first.
1. But what if we get away from the task of "distinguishing" a trend from a flat or a flat from a trend?
2. At least it works like this: if the trend is trending, then the trend "part" of TS earns, while the flat part works in 0 and vice versa, while the flat is profitable, but the trend part does not.
1. To escape (fully and harmoniously) from the task of recognition is precisely the definition of a flattened-combined-trend.
2. The easiest (and most reliable) way to do it is through switching. Besides, in this case the requirement of non-creaminess of both systems is weakened. Which is critically important. Otherwise - to point 1.
On the first point: What I like most about this track is the rejection of the concept of a flat. By replacing it with "balanced alternation of short trends".
1. First you have to find correlations among the strategies. 2. Then trade them through the weights for each strategy.
1. If you can give an example of the relationship.
2. Agreed. What do you think :
- in a neutral situation, in which direction should the centre of gravity be shifted, in a trend or a flat working strategy?
- We state flat, so the user defined his current situation, in which direction should the centre of gravity be shifted, trend or flat strategy?
- trend, which way should the centre of gravity be shifted, trend or flat?
Who is thinking about this? Without being specific, has anyone implemented "combining the incompatible" into a single TS? Not an alternation of one with the other, but a complete and organic combination of a flat strategy and a trend strategy in a single TS... What are the principles and characteristics of such a "hybrid"?
No, you can use weighting coefficients and select them automatically. But still, it is not certain that the maximum of the target function will be higher than zero.
It depends on the implementation. Reshetov has an example lying somewhere in codebase. It is based on three lattices: one flat, one trend, and one toggle.
It works, for better or for worse. // With all the usual restrictions on out of sample, of course.
The interesting thing about this scheme (TrTS+FlTS+Trigger) is that all three parts can be independently tweaked and improved. Which is a definite plus.
1. To escape (fully and harmoniously) from the task of recognition is precisely the definition of a flattened-combined-trend.
2. The easiest (and most reliable) way to do it is through switching. Besides, in this case the requirement of non-creaminess of both systems is weakened. Which is critically important. Otherwise, go to point 1.
On the first point: Most of all in this way I am impressed by the refusal of the notion of flatness. By replacing it with "balanced alternation of short trends".
It depends on the implementation. Reshetov has an example somewhere in codebase. On three lattices: one flat, one trend, and one toggle.
It works, for better or for worse. // With all the usual restrictions on out of sample, of course.
The interesting thing about this scheme (TrTS + Flux + Trigger) is that all three parts can be independently finalized and improved. Which is a definite plus.