
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
...
Certainly the picture is far from complete but by some characters you can
already draw certain conclusions...
...
Niroba, are you trying to play with dolls? Or wolves and sheep :) ?
And in my opinion, there's not much more to look forward to. Because, at least to the one who started this thread, there is absolutely nothing to say. He did put up a new post today, but unfortunately the moderator saw fit to delete it completely. Too bad. This Niroba wrote to me on the third day
I wanted to learn from him, but I didn't get the chance. Apparently this great expert on good manners, not only can not sometimes pick up the words, but when they do pick up, they are not very censorious.
Well, what's the matter with him. The question is different.
If the initiators of the thread, except for words can not provide anything, then what are we doing here?
Why this thread is still alive?
The indicator, on the basis of which I make my forecasts, is freely available on the spider.) I will put it here as well - it is an indicator of Murray levels.
Now a little bit about the mathematics: The main problem with these levels is that it's very easy to indicate their significance a posteriori, but it's non-trivial to define them a priori. On the native site there are a bunch of ways to improve the situation, but they did not seem mathematically sound to me (as well as the theories of Gann and Elliott - that is, the qualitative description is high, but when it comes to quantitative estimates - then the uncertainty).
In a given situation for the solution of a problem (about a statement and methods of the decision once tried to come up on a spider in a branch about a centiment, but they have not considered interesting there :) - it's easier to discuss common themes than to work out a method with quantitative evaluations :) - However that's in the past) yes, about problem statement, postulates and selection and evaluation of suitable solution methods is a separate topic. The most important thing that was done - the problem is set in such a way that it implies the possibility of non-random prediction, as well as some probabilistic evaluation of the obtained results. Now I omit the question about the theory of market efficiency - I want to note that it is only a theory, no more and no better and no worse than other non-self-adversarial approaches. For example R\S statistic (also known as Hurst criterion) gives more than 0.5 for EUR price series (0.64 approximately - that's if you take all series) which in its turn means possibility of non-random prediction and advantage of trend forecasting models. I, for instance, have taken a different approach - there are times when reliable forecasting is possible (Hurst coefficient is significantly different from 0.5) and times when it is impossible (not much different). There are times when trend methods are advantageous (k-factor about 1) and when counter-trend methods are advantageous (k-factor about 0).
The result is as follows - we obtain not only Murray levels from which pivot points are derived, but also their statistical significance at this point in time. That is one and the same level at different times appears in different appearances (how appropriate :) ). That is, at some time this level may work as a breakout one, in some cases it is not taken into account at all and sometimes it is a breakout level, though according to the strategy when the level is defined as a breakout one, a trader should hold a position, this position is usually in profit and the only thing is to move the stop to an appropriate level.
More precisely, as a result of calculations, trader will obtain some turning area limited by Murray levels and channel borders (it is when projections of confidence intervals are built) - the level of confidence interval itself cuts off forecast fulfillment probability levels, while intersections of channel borders and Murray levels provide valuation by time. You can see it clearly on the charts.
This is in general, so to speak "on the fingers".
To be more specific, I supplemented calculations of Murray levels with prediction of a trend movement by a linear regression channel + calculation of Hirst criterion for a given channel (the sampling criteria for drawing the channel are non-trivial too).
Reasoning for the choice: I wrote about Hurst criterion above, and about channels: You can build quite a large number of linear regression channels at any moment in time, but only the linear regression channel will have the minimum error. For forecasting it is better to take the best approximation at a given time. That's about all - that's enough to start with ;) . This will save a very considerable amount of time that has been spent on research. If one is "bored" with programming - almost all this can be obtained using MT4 standard means - except for Murray levels and Hurst criterion, but levels are freely available, while without Hurst they work not much less effectively.
I should also point out that in previous versions there was a small error leading to wrong history drawing. So you'd better get it here:
The interpretation of levels is essential - that's why the part of the article that the indicator is made from has been placed inside as comments.
But changing MMPeriod variable (by default it's 1440 - day) we can display levels of any other period on any timeframe (of course for the proper work the period of displayed levels should not be less than the current one).
Perhaps I have presented my thoughts a bit haphazardly.
Good luck and good trends.
Thanks Vladislav! I feel that finally we have a substantive conversation that is directly relevant to the subject of this forum! It is nice to talk to a person who understands what he is dealing with and how to deal with it (FOREX). I, too, firmly believe that FOREX should be communicated only on the basis of statistical data, which can represent numerical characteristics. Because all methods that give a qualitative estimate, which include, among others, zigzags and Elliot's methods, is a game of luck, which must be played while staying near the monitor almost continuously. Basically, those who have learned to play it, are very good. But most people can not learn to play it. And it's not even in the amount of time spent studying the game.
I would like to clarify on the posted indicator.
When compiling it says that variables mm_period, StpBck are not defined. Can you specify the values of these variables?
The code was corrected - please re-download it. The fragment was inserted from another program :). Everything compiles now.
One more thing I forgot - when setting MMRead = 0 you'll see the corresponding octave (its dimension is specified by P variable and 64 by estimates is the best option) built by data of current tf.
And the most important thing: this approach works not only on Forex instruments which gives hope for statistical validity and not too fast crash (and most likely the method can "remove" latent inefficiency of market movement and the crash is not expected, though for sure the method is not the only one ;) ), but to me as a person with more or less appropriate mathematical education it seems reliable.
Good luck and good trends.
Good luck!
I'm very interested in the part of your strategy where "reliable prediction" of trend and counter-trend methods of operation takes place. Can you tell us more about it? In other words, how do you predict that the market is in a flat position and will remain in it for some time?
I cannot make prognoses. In my strategy I simply divide the market phases conditionally into two. Phase 1 - any activity in the market (strong trend, news, etc.) and Phase 2 - the calm phase (sideways channel). I do it using a trivial method - the Standard Deviation indicator, which is part of the MT4 delivery. I simply define that such and such deviation period and indicator values will show me these two phases and on this basis I either place limit orders (if we are currently in flat) or simply remove them (if there is some movement, i.e. indicator values have exceeded the threshold). But it should be noted that this indicator only shows the current state of the market! It cannot show you or predict what will happen in the next hour, for example! And you, making a statement about the possibility of "reliable prediction", how can you prove it? Please provide a detailed description of the forecasting methodology you use.