To follow up - page 10
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
Walking away from the MAs, why? Because, apparently, you didn't like the result. Have you ever wondered why they do that? Because they try to filter out what has a very distorted relationship to market movements. And with simple-mindedness as an argument, nothing else can work. Try reversing it in your head by putting the micro-purpose - i.e. the minimum trading opportunity (micro-context) - at the core of the discretisation. Not time. Well, even from general observations it is clear that there is no special point in filtering by bars of the timeframe. Here we have normal volatility, active trading - everything is OK: large fluctuations consisting of small ones that are practically filtered by BP. Everything is OK. But the movement is gone. I.e. it is gone when I use time as the X-axis. All this filtering is useless. Inventing all sorts of MAMA-HAMA, Yuriks and a bunch of other stuff that tries to adjust the phase/period of the filter to the changed conditions. I.e. to use micro-context in one way or another.
But that's the approach through the tonsils, as they call it!!!))
Filter a signal that already initially contains samples of that micro-context. That's what I'm talking about...
===
Well try looking at the market not as a series of discrete time intervals, but as a series of discrete states.
Well, try looking at the market not as a series of discrete time intervals, but as a series of discrete states.
>> Odds? Odd? >> Rib?
Sorento, составь бары по сигналам пересечения машек. Наверно, на это Петр и намекает.
And what's the discreteness? Smoothing parameters? ;)
Sorento, составь бары по сигналам пересечения машек. Наверно, на это Петр и намекает.
You could do that, but you don't have to. For now, the main thing is to understand the transition from time-based to context-based counting. And what context will be highlighted and how... In short, let's first settle on a general approach.
Walking away from the MAs, why? Because, apparently, you didn't like the result. Have you ever wondered why they do that? Because they try to filter out what has a very distorted relationship to market movements. And with simple-mindedness as an argument, nothing else can work. Try reversing it in your head by putting the micro-purpose - i.e. the minimum trading opportunity (micro-context) - at the core of the discretisation. Not time. Well, even by general observation it is clear that there is no special point in filtering by bars of the timeframe. Here we have normal volatility, active trading - everything is OK: large fluctuations consisting of small ones that are practically filtered by BP. Everything is OK. But the movement is gone. I.e. it is gone when I use time as the X-axis. All this filtering is useless. Inventing all sorts of MAMA-HAMA, Yuriks and a bunch of other stuff that tries to adjust the phase/period of the filter to the changed conditions. I.e. to use micro-context in one way or another.
But that's the approach through the tonsils, as they call it!!!))
Filter a signal that already initially contains samples of that microcontext. That's what I'm talking about...
===
Well try looking at the market not as a series of discrete time intervals, but as a series of discrete states.
Isn't it easier to forget about all sorts of discretizations and indicators and not to look for trading system ideas on indicators and other price derivatives, but to start from the zero level of price-time-volume and try to formalize the found features. Perhaps, we will formalize them with the existing indicators (since there are not so many variants and almost all of them have already been implemented in the form of an indicator), but without excessiveness and with the full understanding of the logic. And to investigate not the indicators behavior, but the initial hypothesis of the price behavior and to return to it constantly if changes are needed?
Then it doesn't matter whether the chart is presented in zigzags or bars. If amplitude modulation formalization is needed to describe the next hypothesis about the price behavior, we can try ZZ as well. Not bound to anything and not limiting ourselves with formalizations from the very beginning because it limits the range of price behavior patterns that can be found
Isn't it easier to forget about all kinds of sampling and indicators and not to search for trading system ideas based on the indicators and other price derivatives.
It is not clear. What could be more logical than to formalise the trading context? It is the 'feature' that I formalise. The attributes of the flow cat. themselves are sphere.tabun.
Maybe we will formalise with the existing indicators (as there are not many variants and almost all of them have already been implemented by someone in the form of an indicator), but without being exquisite, fully understanding the logic. And to investigate not the indicators behavior, but the initial hypothesis of the price behavior and to return to it constantly if changes are needed?
Yes, I think that's what I am trying to do. The indicators are the same. The subject of measurement is different. And using indicators without understanding their logic... Ha! I certainly don't need convincing that it's... I can't even find a decent word for it.
And the context approach precisely exploits, among other things, the "original hypothesis of price behaviour" as an argument for analysis. I probably need more detail than I've already written. ok. I'll formulate it and set it out.
Then it doesn't matter whether the chart is presented in zigzags or bars. If we need amplitude modulation formalization to describe the next hypothesis of price behavior, we can try ZZ as well. Not getting attached to anything and not limiting ourselves with formalizations from the very beginning because it limits the range of regularities in price behavior which can be found
So I'm not getting attached. On the contrary, if you noticed, at this stage of discussion I avoid any specifics, any methods of formalization. Until an idea of the proposed approach is settled, to go on with specifics (which - you're right - can be anything, why limit yourself?) - means to get bogged down in irrelevant details.
Okay. In follow-up to xms. Somehow spontaneously recently in some thread advised someone (for the cause!!!)) to learn a ditty to the tune of Jingle Bells - the words just came together in my head, it happens:
Margin calls, margin calls,
Margin all the way!
Oh what fun it is to ride
In a financial sleigh!
This is actually the chorus. Then I tried to think of something for the verse, made a few variations, but, as it often happens without spontaneity (inspiration?)), the result was not that. Rejected.
Maybe someone will continue. Let's play a game of bloody burime...)))
And what is the discreteness? In the smoothing parameters? ;)
I do not understand the question. We are not looking for discreteness, but first and foremost visibility. With this conversion of the original time sampling to a "pseudo-signal" sampling in front of us, we can now apply to this representation what was originally intended to be a filter. But now this 'filter' becomes an indicator in its own right, superimposed on the new graph.
Our ideal task is to make such a transformation from the very beginning so that nothing else has to be applied to it, i.e. that the system would already be immediately visible as profitable.
Martin all the boy!