a trading strategy based on Elliott Wave Theory - page 217

 
Yurixx 13.01.07 00:12
...
By the third parameter did you mean the value of the price change or the success of the trade ?

The third parameter is what you are interested in. At this stage, I don't think it's possible to determine what exactly should be compared, so you'll have to look at both.
 
Yurixx 13.01.07 00:12
...
Под третим параметром Вы подразумевали значение изменения цены или успешность сделки ?

The third parameter is what you are interested in. At this stage, I don't think it's possible to define exactly what you want to match, so you'll have to look at both.


I'm interested, alas, not even in the number, but in the function. Exactly, the dependence of the probability of a successful deal on Take Profit and the value of Stop Loss that should be used in this case.

But we can, as a first approximation, fix an admissible minimum probability of success at which the EA opens a position and then just look for the TP level, i.e. the minimum change of price in the necessary direction at a given probability.
 
Yurixx 13.01.07 02:24
...
I'm interested, alas, not even in the number, but in the function. Namely, the dependence of the probability of a successful trade on Take Profit and the value of Stop Loss to be used.

But we can, as a first approximation, fix an admissible minimum probability of success at which the EA opens a position and then simply search for the level of takeprofit, i.e. the minimum change of price in the necessary direction at a given probability.

The method of enlarging the parameters and observing the result, otherwise there is a great probability of getting bogged down in analytics (the solution is often not guaranteed). But be prepared to be accused of "over-optimising" the system.
 
Например, так ли уж важно знать значение изменения цены? Если да, то может пренебречь самими индикаторами:-)


In fact, the value of the price change is irrelevant. :-)) It's not a pun. It's just that it (and only it) can be used to build a criterion for selecting successful trades, and hence calculate their probability. If I'm wrong, what is it ?


Now, put this trading algorithm into a tester and trade each indicator value, assigning it a 1 if the trade brought a profit, and -1 if the loss. Next, build the same area on the same plane, but paint the points red and black (for example), respectively, depending on the sign. If one point of the area has several identical indicator readings, you should first sum up their "colours" (+1 and -1) and, depending on the sign of the resulting sum, determine the final colour. I advise to remove the commission for each trade in the tester at this stage, it will allow to estimate the maximum potential strength of the strategy.
Thus, the area will be divided into three sub-areas:
positive, negative and mixed trading results. We will choose the one we want. Define the borders of the area, and so on.
Write down the result. We will discuss.
 
2 Yurixx
For reasons unknown to me, the download was incomplete, so I'll probably finish it tonight.
I will probably download it tonight.
 
MathCAD cannot be uploaded to the website. At the end of the upload it gives a list of errors. Some kind of xml not found. :о(
 
A little off-topic for the current topic. Here's one of the typical mistakes (I mean, my prediction also sometimes lies) of the prediction algorithm. A little bit wrong with "skylining" existing fractals, but I'm continuing my research.


Yuri, in some ways I stopped building phase planes a long time ago (by the way, they are not phase planes at all). Including building on bulls and bears. But I hope you will be able to realize your TS on such approaches.

Note on the current topic, it seems that Sergey has published his studies and, in my opinion, got the correct conclusions concerning the optimum number of indicators and, most importantly, that indicators should be uncorrelated. And are you sure that the bulls and bears are? All you have to do is look at them:


Or am I wrong?
 
Here's one of the typical mistakes (I mean, my prediction also sometimes lies) of the prediction algorithm. A little bit wrong with "skylining" existing fractals, but I continue my research. <br / translate="no">.

Or maybe it's not an error, but just a manifestation of the fractal nature of the market? That is, at the point where the divergence occurred on the chart, the support was simply broken (stops triggered) and the market followed an alternative scenario? That is, based on previous data it was simply impossible to assume the second scenario? I think that a trading strategy should be able to consider such possible market development scenarios. For example, in such cases I use trailing stops that return about half of losses from stops. From my observations if a stop is rationally chosen then price after its breach in most cases is able to go some distance back, due to which there is an effect of some compensation for losses from an unsuccessful (premature) entry into a position. I am now collecting more accurate statistics on this statement by observing the work of the Expert Advisor.
 
Вот одна из типичных ошибок (это я к тому, что мой прогноз так же иногда врет) алгоритма прогнозирования. Немного ошибается «скайлинг» существующих фракталов, но продолжаю свои исследования.

Or maybe it's not an error, but just a manifestation of market fractality? That is, at the point where the divergence occurred on the chart, support was simply broken (stops triggered) and the market followed an alternative scenario? That is, based on previous data it was simply impossible to assume the second scenario? I think that a trading strategy should be able to consider such possible market development scenarios. For example, in such cases I use trailing stops that return about half of losses from stops. From my observations if a stop is rationally chosen then price after its breach in most cases is able to go some distance back, due to which there is an effect of some compensation for losses from an unsuccessful (premature) entry into a position. I am now collecting more accurate statistics on this statement by observing the work of my Expert Advisor.

This is an interesting idea. I did not even think of alternativeness. The final result of model processing I intended only for calculation of channels (there are none yet, but only a very approximate estimate of price values (or rather extension of fractals on the price plane) for creation of channels), moreover, at "macro level". I was not interested in all alternatives of "small forms" but I had not even thought of large forms. The most interesting thing is that my model will theoretically be able to produce such alternatives. Solandr, thanks, something to think about...
 
2 grasn
I note on the current topic that Sergei seems to have published his research, and in my opinion, got the right conclusions on the optimal number of indicators, and most importantly, that the indicators should be uncorrelated. Are you sure the bulls and bears are? You just have to look at them:<br / translate="no">

Sergey, I don't use any standard indicators. Not that it's a principle, but it seems to me that what lies on the surface and is known to everyone cannot be a source of fabulous wealth. :-)) Especially it cannot give the joy of creativity or scientific (okay, quasi-scientific) pursuit. And what else, apart from these two things, can attract in forex ?