Market phenomena - page 17

New comment
 
paukas:
Slava, can you check the hourlies, are there any spikes at 12 pips?

Vladimir, at 10 pips there is a spike (maximum) and 12 is right after it
 

paukas:

...so that even the rank-and-file Ktn would understand.

I'm sorry, I just want to check.


Are you a candidate? I take it you're already in some kind of state of mind? I'm afraid it's not that simple. It'll take me a while to become a candidate as well. I will be - you'll see. And once I'm a candidate, you and I will start interacting better. Can you imagine the life from now on? No? Neither can I, but I want to try.

Wait, take your time. Becoming a candidate is a complicated process that will take some time. Don't go anywhere, I'll be back as a candidate when the urge to hit you with something heavy passes.

 
Farnsworth, why is there no symmetry in your table with respect to 0. The omega process for negative values starts at -0.0022... and less, and in the positive zone 0.0024... and more. So negative omega increments will be bigger as the range is wider, so by itself omega has a downward drift, and alpha has an upward drift? It is the same when you go over 0 - alpha has a wider range in the positive zone than in the negative zone. You have constructed the matrix in such a way that one "process" knowingly has more negative values and the other has more positive values and got what you got - a downward drift of one and an upward drift of the other
 
Avals:

Vladimir, at 10 pips there is a spike (maximum), and 12 pips immediately after that.
10 pips is a good figure. people do not bother to calculate stops accurately, they put plus or minus ten - but what the hell!
 
Farnsworth: ....HideYourRichess, ... No need to wring your hands and faint, not that I want to offend, but I know how histograms are constructed without you (of course respected ones). Sometimes you just have to look at things a bit outside the box. That's all.

I am surprised at your unwillingness to read.

Well, there's no explanation for these holes. Whereas it seems clear where the bumps come from, it's hard to think of holes other than excessive grinding of the intervals.
 

Indeed, it's hard to believe that everyone is fixated on these histograms. There are plenty of examples when the phenomena were discovered due to some incorrectness, such as unwashed test tubes).

All the more a more sympathetic recipe for a division into processes has been given, although not in sufficient detail. And the more so that the idea of division into two processes has been floating around on this forum for some time, for example avtomat (not to go far) has also got a mixture of two processes, the scheme is not quite clear though.

I have a question to Sergei, in the second recipe the same rows are allocated or not quite the same?

 
Avals:

Vladimir, at 10 pips a spike (maximum) and 12 just after it
Thanks, that's the way it is. You can get about half a pip per trade from that.
 
HideYourRichess:
Well, there is no explanation, these holes. If for bumps it seems clear, where they can arise from, then for holes, except for excessive grinding of intervals - it is difficult to think up.

What the hell with the holes). Especially if you want, you can come up with an explanation for everything.) For example, a row of "recommended sizes" for one process is different from such a row for another ).

 
Candid:

What the hell with the holes). Especially if you want, you can come up with an explanation for everything.) For example, a number of "recommended sizes" for one process is different from another.)

In fact, it is not about making up explanations. In the same example with cones, there the explanations are not invented, but practically testimonies of eyewitnesses and the authors of these cones.


In other words, I am not interested in the holes themselves, but in what is in them, provided that there are holes after all. Otherwise, how to enter the market if you don't understand what is going on. I do not want to go far and if I've understood correctly, the machine has another approach, namely the description of the process. True, I think that there are not two of them, but a variable number. But it all depends on the level of generalization.
 
Avals:
Farnsworth, why is there no symmetry in your table with respect to 0. The omega process for negative values starts at -0.0022... and less, and in the positive zone 0.0024... and more. I.e. the negative increments of the omega will be bigger as the range is wider, so by itself the omega has a downward drift and the alpha has an upward one?

"thought" is ahead of my hands, I mean, I did make a few mistakes, which probably led to questions and misunderstandings. I did not explain in detail the primary objectives of the study. They were related to the following areas:

  • To study the impact of "fat tails" and outliers in the incremental flow on the price trajectory
  • studying the probability for certain events, such as occurrence of certain classes of trajectory deviations (i.e. price deviation from the primary level) in certain areas
  • Search for an adequate model for a dynamic system with random structures - particularly search and classification of systems and their parameters.

As to the first direction, I probably should have started with it and put another phenomenon (I have already described it briefly with apologies). It is very interesting, but from a practical point of view, it seems to me, not very useful; a trend can be identified, but there are serious problems with

  • with modelling this trend, a very complicated process, this "bearing" process cannot simply be modelled
  • with predicting a killer process, and that's not even practically possible. The "killer process" is the second sub-process that destroys the trend and the result is what we observe.

As to your question, this is a way of classifying and accurate methodology. You can divide the increments into positive and negative. There will be two processes - only positive quotes and only negative quotes. Would that help trading? I think not. The classification you have laid out is the result of more complex analysis and isolation of these very "random structures", including after studying the above. You can probably do it without inverted commas. The purpose of classification is to make sub-processes more predictable and transitions between them usable in practice.

If this direction is interesting, you can adjust this shift and see the result, nothing will change in principle. The shapes of the subflows will remain the same, they will just be "crooked".

PS: if it is not clear again, I'm sorry, I can't explain otherwise.

1...101112131415161718192021222324...75
New comment