![MQL5 - Language of trade strategies built-in the MetaTrader 5 client terminal](https://c.mql5.com/i/registerlandings/logo-2.png)
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 initially misunderstood and responded too hastily. Then I erased the post when I saw Peter's reply.
All right, jokes aside. I'm not going to guess which way Benya will fart, but I don't need to. Wherever he farts - the market doesn't care about it, and he will do it his way anyway (examples abound: for the terrorist attack of 09/11 Masterforex showed it convincingly). Or, Ben simply has nowhere else to go and is forced to fart where the market makers tell him to fart in his ear. But in almost any case, the market will still use Beni's farting as an excuse to play with liquidity.
And that at some point there is bound to be an event equivalent to a micro-panic in Dubai, the market has long known.
Peter, all your intricate reasoning about contexts, chains of microcontexts, clustering by criteria still ultimately boils down to one short phrase: "you can go in there" or "sit on the fence". By doing so, you make a short-term prognosis, to put it another way: "the current situation is such that statistically this context has a continuation".
But the main point is that for a purely martingale process and given the information only about it itself, this reasoning makes no sense, because there is no continuation of the context for it. At any given moment it can suddenly cut off and change to the opposite or neutral. But the reality is different: for you, given your practical experience, the context does continue more often than it breaks off, i.e. the process is not martingale. Consequently, it has a memory. And memory in turn implies that at least sometimes the process has information about what it will do next in the short term.
I continue to believe that any trader who systematically beats the market over a long period of time - whether he knows it or not - still makes predictions and actually takes advantage of the non-martingale nature of the market. (I would like to point out, for the natives of Neobit Island, that Pastukhov also assumes that the market is a non-martingale. It would be highly suspicious if a mechmate defended his dissertation on a profitable martingale playing system).
Alexey, just super ! And on such a cool emotion. :-) I think this is the most rational understanding of the market process. General enough to finally stop chewing over the terms prediction, memory or inertia of the market, its martingale, etc.. And specific enough to soberly approach the creation of a TS.
I don't understand why there is suddenly so much debate around the concept of "context". Firstly, the concept itself is so general that you can shove almost anything you want into it. Secondly, the author has not given any definition and thus has not restricted freedom in any way. Thirdly, it fits perfectly with the above-mentioned notion of the market process.
I do not understand why there is so much talk about time. There are no dogmatic fans of astronomical time and no frostbitten opponents of time in general. But there are arguments. It would seem a sensible idea that a specific form of time use should not destroy useful information has already been stated here implicitly. What may be opposed to it? If this form cuts off information not necessary for this TS, it is a problem not of the attitude to time but of the adequacy of this TS.
In this regard, isn't it time for the high assembly to move on from general talk to something more concrete.
For example. Talking about context in general, will never lead to any result until the circumstances that define that context are identified. Even better, the values that parameterise it will be named. Then it will become clear whether this context is indeed capable of expressing market conditions and providing some means of making a profit or not. At the very least the debate will become more meaningful.
The same goes for the time. In order to get beyond what has been said and repeated many times, we need to discuss specific approaches. Because there is no question of "is or isn't" but a question of what time and how it is used in this approach, i.e. it helps or hinders its meaningfulness and adequacy.
Uh-huh. That's why I was so reluctant to start with a general reasoning. "Field rotor like divergence..."))
===
Lyosha, the passage quoted by Yurixx is the one you erased and, as I understand it, my reflection is not required? As if the question is removed?
Something I do not have such an impression. But... whatever.
No, I didn't erase it, I just erased Candid's wrong answer about clustering.
Who's stopping you from reflecting in your own thread? The question is... well, yes, it's off - in the sense that our positions remain unchanged. But you have not become my enemy because of it :)
Better yet, post your ranks if you're cured (I'm not talking about that drugstore). I'm looking forward to it. It'll finally clarify what you mean by context. Yuri didn't bring it up again for nothing.
Well, yes, it would be more concrete if you had a concrete example :)
I actually wanted to clarify with my question whether an implicit context assignment is allowed. Since I interpret my example as an implicit assignment.
However, it seems we should do the same thing as in the case of an explicit assignment, namely: learn to recognize at an early stage the emergence of a useful context (in both cases given to us in the examples). Similarly, we will need to recognise early signs of the onset of an unfavourable context (or the end of a favourable context). But if this is the case, why not do it by a temporal CR? Although if we want to predict its occurrence, a sequence of antecedent contexts, i.e. a frame representation, might also help us.
2 Alexei: I didn't quite understand where and what you erased, but since there is such a reason, I also erased my post. Just in case. :)
No, I didn't erase it, I just erased Candid's wrong answer about clustering.
Who's stopping you from reflecting in your own thread? The question... well, yes, removed - in the sense that our positions remain unchanged. But you have not become my enemy because of it :)
Better yet, post your ranks if you're cured (I'm not talking about that drugstore). I'm looking forward to it. It'll finally clarify what you mean by context. Yuri didn't bring it up again for nothing.
Ah, yes - sorry. I did read it carefully. (Yeah, there really is something wrong in the native Elsinore)).
That's the thing, inertia (here, the continuing effect of the original causes) by RRR (or whatever) is hard to see in TF. I mean, you can see it, but it's a matter of luck. So... Fragmentally, when the stars align, or to be more precise, when the timeframe coincides (exactly or multiply, by harmonics) with, say, the volatility frame.
People keep trying to prove to me that there is no inertia, arguing either with the absurd "the market is not physics" (am I arguing? although Newton's 1st law fits quite well as an analogy), or with the help of ingenuous statistical deductions for BP.
I say, take the same mountain canal built on ZZ, to avoid going far and complicating things, draw its median, take the resulting series, where the elements will be its - medians - horizontal sections, and count from it your favorite PP, etc.)) Feel the difference! Nope. Futile - the cactus tastes better.
===
Rows to post... I'm wondering what's the best way to do it. Apparently, the formed series should be merged into a file, which can then already be analyzed in Matlab, Excel or wherever else.
OK. Or what, any thoughts?
That's the thing, inertia (here, the continuing effect of the original causes) on the RRR (or whatever) of the RRR in tf is hard to see.
[...]Rows to post... I'm thinking of the best way to do it. Apparently, the formed series should be merged into a file, which can then already be in Matlab, Exel or wherever else to analyze.
Well, who said it would be easy. The finance giants would not be interested at all in bringing Forex to the masses if the truth (=inertia of reasons) lay on the surface.
And put the rows in csv format. I hope this format is not only readable by Excel.
Well, yes, it would be more concrete if you had a concrete example :)
I actually wanted to clarify with my question, whether an implicit setting of the context is allowed. Since I interpret my example as an implicit assignment.
However, it would seem that next we should do the same thing as in the case of an explicit assignment, namely: learn to recognize at an early stage the emergence of a useful context (in both cases given to us in the examples). Similarly, we will need to recognise early signs of the onset of an unfavourable context (or the end of a favourable context). But if this is the case, why not do it by a temporal CR? Although if we want to predict its occurrence, a sequence of antecedent contexts, i.e. a frame representation, can also help us.
I guess you don't mean predicting, but recognising its presence or expiry? Although I guess I could do some predicting, but that's not with me. Nostradubus is a dime a dozen without me. The most visited branches are for all tastes. Predicting the context, i.e. the situation resulting from something that has influenced the market, is tantamount to knowing what and with what consequences will happen in the future (with what you can, but "what" is out of the question). Impossible. And even if it is possible, the reliability of prediction and recognition are two big differences, and therefore, from a practical point of view, the former is of little interest.
Why just frame by microcontext may be better to solve this problem than by time? Well because nobody cancelled fractal-design of the market, and the big is reflected in the small. Each large context has its own characteristic micro-context. I will certainly show this later. I have already given the example of adaptive AI, which tries to put a tailcoat on the context, but as they are initially, first of all, tied to time, the "tailoring of the tailcoat" is limited by it. And fuck knows, how long will last, for example, the dead calm - may be ACs trade with themselves as undermined on a "small", and in the case of the exchange - market makers (here as techn. f.) hold in the absence of market ideas corridor and liquidity. It has nothing to do with recognition (here - within the trend model).
===
In general, you are right - it seems it is time to stop with the "context" quibble. Especially as I seem to have managed to get the general points across. It is time to move on to specifics. But I reserve the right to ramble on for a while. Don't mind me.))
I guess you didn't mean to predict, but to recognise its presence or expiration?
Yes, that's what it said, "recognise". And predict as an alternative. Because this is where we started discussing wrapping frames into specific bars. I think that such a collapsing would mean transition from recognition to prediction, because for recognition we need to be inside the frame, don't we? And once it's convolved, we can't see the inside...
And microcontexts are patterns? Patterns are any recurring states that extend over time.
As far as I understand Peter, the very notion of context is already linked to prediction. Implicitly.
Peter, I hope I'm not stepping on your toes with this statement. I am not going to get into a tug of war over whether we are making predictions (or forecasts) or not, I am not interested in that. However, every context is necessarily (!!!) linked to an attitude to buy, sell or sit on the fence. And in that sense it is in itself already an implicit prediction of future market behaviour. Another thing is that it can be reversed in the next second, even if it occurred just a second ago. Isn't it so? But that's what any TS goes under, because it's the market.
So, imho, let's just leave this prediction fight alone, it's not getting us anywhere anyway. And also the issue of predicting context, as being deliberately tautological.