Hidden divergence - page 35
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
Well, I'm not trying to discourage it. Simply, if I have a possibility of precise formula analysis, I use it and try to find out "why", i.e. the logical reasons. If I don't have such an exact opportunity, I try to perform statistical tests on as much material as possible. If the stat tests steadily tell us that we have got the cat by the tail (caught the stat advantage), then I am happy with this explanation as well, i.e. just the "how".
Well, who needs it, he understands all hidden "attacks" in his address .......... and for "who?", and "for what?" - you'll get tired of answering questions :)
There is such a picture, it is not one, of course, but for an example this one will also do......... "hidden"...... without filtering:
.... oops - third time I can't attach a screenshot :( ... beats the pride of a sysadmin :)
attached... in rar :)
I can't make a screenshot, the only question is, - is it possible to get pips in + points from this case or not?
to Mathemat ..... now I'm mastering the statistics, it used to be written to a file, now it is not :(.... I messed up the cycles somewhere.... but it's fixable.... I'll do it.....
the question is: for me it's a "dark forest", all I can come up with is a dumb entry and exit by taku or stop.... drawing of course :)).... but if there are wishes of any kind, then I will try to implement..... so that "statistical tests steadily tell (tell) us":)...... because, already without drawing, - statanalysis really is a "dark forest".
s2101 has recommended parameters 9,21,5 in the OsMA oscillator. Look at the picture. Oscillator with these parameters, unlike oscillators with normal parameters,
it gives a false signal. I know in advance what he will say that we should look for other timeframes, other indicators and wave analysis. However, I have concluded for myself that it is better to use the original OsMA parameters. Let the oscillator give a signal later than a false signal that misleads. And then, these parameters have been verified by many years of historical experience.
At the beginning of this thread (or discussion in general) there was (and is) a certain Rosh, who is known to all (one of moderators) and defined it all with one phrase (not a quote): don't look for a leading, look for a good lagging..... =)
Yesterday's entry on the pound
100п
not on divergence and not on convergence
---
Take profit exit
---
Obviously, it is good to exit on one of the signals, but the problem is two signals!
---
it was possible to set the semi-automatic to exit by the signal, but it is obvious that the loss of the profit by the first signal is half the profit taken by the take
but the second signal has a higher profit margin.
----
of course latent and non-hidden diverters are good, but they should be mixed with other products
I mean, you don't just need a single signal from some divergence indicator
you need a complex analysis
---
yesterday's entry and exit was not based on a divert
but by the general movement of the market
---
obviously the output is good on one of the signals but the trouble is there are two signals!
---
which one was it before? ;)
but what is it before?
I marked them both with circles.
The second is certainly better than the first.
But I put the TP, it's more reliable.
the exit signal first circle - formed when the profit was about 50p
second one is when the profit would have been over 100p
----
in theory if an entry is made the exit can be instructed to a semiautomate
i.e., a program designed for an exit
which will close the order on certain signals - whether I will sleep or take a walk or rest
I gave an example of how, in this case, closing by the first signal gives a big loss - almost half the profit
and the second signal increases profitability
the frequency of such signals is not good quality
I marked them both with circles.
the second of course better than the first
But I put the TP, it's more reliable.
the exit signal first circle - formed when the profit was about 50p
second one is when the profit would have been over 100p
----
in theory if an entry is made the exit can be instructed to a semiautomate
i.e., a program designed for an exit
which will close the order on certain signals - whether I will sleep or take a walk or rest
I gave an example of how, in this case, closing by the first signal gives a big loss - almost half the profit
and the second signal increases profitability
the frequency of such signals is not good quality
sorry, but I was asking as a "tease"..... :( you may be way ahead of us all here, but in this discussion the question of position maintenance has not yet been asked.... the entrances are still trying to suck in from all sides :))
i'm sorry, but i was just teasing ..... :( You may be way ahead of us all here, but in this discussion the question of position management has not yet been asked.... entrances while we are trying to suck in from all sides :))
I don't think entry and exit should be considered separately! if you are trading on a system I guess you have to decide how to exit at once
of course it is better to exit by the return signal plus stop and trawl options
this is where the problem arises because there are a lot of signals! and it is necessary to apply some kind of filter - otherwise nothing will work
---
Here's another interesting thing
when i was writing - or rather testing - my Expert Advisor of 2007 (which worked on divergences)
there were a lot of false signals - they would have annihilated all my efforts
I came to the conclusion that it is necessary to withdraw to Breakeven early enough, because the profit on this signal usually comes quickly
if i got a good signal, as a rule, the order flew quickly to breakeven
but if we got a signal, we managed to hold a profitable position for a long time by using a filter and some subtleties in working with orders
i used to close entire positions at a reverse signal if the profit did not exceed 50 points
i.e. i closed part of the positions if the profit exceeded 50p, i.e. the first reverse signal was evaluated purely mechanically
thereby trades were sometimes held up to 300-600 points
of course 50 points is conditional and the selected period parameter is 2007
in 2008 it needs to be adjusted or maybe to change the logic of the mechanics - last year the average EUR move was 60p, this year it is much higher
my filter was 89sma on m15, if the buy divergence was below it i would enter - if it was above i would not go in
if sell diver was higher than sma89 I would enter, if lower I would not go
you may use a neuron net as a filter - it gives a direction signal
----
the main thing is not to use all signals
you need an additional filter criterion - the signals are good but not all of them
that's why i gave the picture as an example
---
one more thing - oscillator is based on averages and lags
Some of the signals you can see on the pictures in the history, in real time they appear too late and in real time some not bad signals cannot be used.
If we set parameters 9 21 5 and any other parameters, on which divergences become more sensitive, then we face the problem of filtering the left signals
Yuri. Looking at the figure, the entry was sort of not on convergence at all (exit is on standard divergence)
How would you comment on the entry (shown by the arrow)
unambiguously "YES" .... Well, maybe, except for one thing - it is worth to filter out worthwhile entries, and then select accompaniment for them.... Sometimes entrances appear "pips" (with almost always minimal losses) and sometimes they are breathtaking ?
Yes, my first thought is "break-even or minimum profit" - but I want more :)
Yuri. Looking at the figure, the entry was sort of not by convergence at all (exit by standard divergence)
How do you comment on the entry (arrow)
Please look at the picture message where I wrote an entry not on a divert - exit also not on a divert
( entry along the entire market segment - breakdown of levels )
I think that not all bears jumped out so fast, some are waiting for a possible move down
and i paused at 100p a week quite normally a fortnight ago it was a bit over 100p
my trading strategy is slow, I don't often buy and sell - my targets are 50 pips or more - I may go over the divider if I swing at all of them - I'm bound to fail.
exit by hand - in the picture without yellow circle - it means exit by hand
i traded in red sell - turquoise buy
---
i was just trying to figure out the best way to exit a trade
and i think it's also clear to you that you don't have to exit by any signal ...
because there are a lot of them and there are just as many bad signals as there are good ones.
as always there is a problem of determining the trend - filtering signals