Strategic foresight systems - page 20
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
how are the results on lower ff? (if anything was done - attach a screen for interest)
I tried it a couple of times, the result is bad, after refinement I will try to show screenshots from lower timeframes as well. I will try to do it with confidence limits.
I see...
well, it will be interesting to see... ( if technically feasible not too long, then on a few sections)
http://www.finam.ru/analysis/macroevent/default.asp?str=2&ind=591×tep=1&dind=0
http://www.alpari.ru/ru/calendar/
All in all, not too bad for a first "mix" (I mean trend change predictions). Well... it's like this, if you don't praise yourself, then... I've got a "finesse" with EURUSD, although according to my classifier it hasn't changed its trend yet. Parallel to the main system I use Markov chains (MC) based prediction, but in a somewhat simplified form (not a mathematician). I dream to combine these models, but I don't know how yet :o( Statistics of change of duration and spread of trends allowed to collect a transition matrix from state to state. For the current (initial) state and on the basis of a known relation I calculate the probability of a future state. The initial state for each new bar is different (there is a completed trend + a "known" length of the current trend that changes). To give an example, Russian roulette, where each pull of the trigger changes the probability state of the "moSk-revolver" system. So, the MC showed an imminent end to the trend, while the main one showed a continuation. Both appeared to be formally correct, but for the time being their usefulness for trading is doubtful.
I've got the first results of cumulative prediction error for 5 future bars over a long period. Here is an example for AUDJPY. Was hoping for a better result. It is not important what "Y" is measured in, what is important is that 0.003-0.004 is the error level of 1-3 days of trend change detection. the error is highly correlated and it is not clear at all what influenced the system identification 200 trading days ago.
We can make some preliminary conclusions:
I am an irreconcilable opponent of TA (I have been for a long time)
Maybe you shouldn't. With TA, with some training, you can be a jeweller or a virtuoso trader. 100 profitable trades in a row is not a problem, and with random inputs even thirty will be unattainable.
No, you cannot. It's a fact and training has nothing to do with it. TA is not a business, but intuition + intuition + intuition + intuition. All TA tools have nothing to do with the market, do not refer to it, do not say anything about it, do not describe it - all these fans, channels, patterns, waves, etc., etc. - is "tuning" of your "brain" and this very intuition. Business is inherently "duplicable", if it's not - it's already art, or just pure play for the sake of play itself.
But this is my IMHO. :о)
No, you can't. It's a fact and training has nothing to do with it. TA is not a business, but intuition+intuition+intuition+...+intuition. All TA tools have nothing to do with the market, do not relate to it, do not say anything about it, do not describe it - all these fans, channels, patterns, waves, etc., etc. - is a "tuning" of your "brain" and this very intuition. Business is inherently "duplicable", if it's not - it's already art, or just pure play for the sake of play itself.
But this is my IMHO. :о)
The same can be said about alternative views :)) But we have gathered for a different purpose, which are no more specific and unprovable at this point in time. Our assumption about the development of the currency market situation can change significantly of course not the swing of a moth, but the specifically occurring and some events not yet known to us.
BUT. We assume that the course of events will not critically go beyond the established ones, and our prediction is based on that.
Looking at the picture it is possible to assume that currencies will get inertia and there will be such tendency on the weeks . But again, this is just a supposition, which by the way is not deprived of one hundred percent logical conclusion.
You cannot believe anybody, not even yourself. One can make an assumption.
Do not judge, gentlemen, that the 0th bar is taken into account, otherwise the indicator does not work.
No, you can't. It's a fact and training has nothing to do with it. TA is not a business, but intuition+intuition+intuition+...+intuition. All TA tools have nothing to do with the market, do not relate to it, do not say anything about it, do not describe it - all these fans, channels, patterns, waves, etc., etc. - is a "tuning" of your "brain" and this very intuition. Business is inherently "duplicable", if it's not - it's already art, or just pure play for the sake of play itself.
But this is my IMHO. :о)
You don't use TA either...
but revelations
;)
The same can be said about alternative views:))) But we have gathered for a different purpose, which are no more concrete and unprovable at this point in time. Our assumption about the development of the currency market situation can change significantly of course not the swing of a moth, but the specifically occurring and some events not yet known to us.
I am not trying to "preach" and I am not trying to agitate anyone. Everyone will find his own way, the main thing is that this way should not become an illusion/ mirage, and it often happens, ..... very often :o(
BUT. We assume that the course of events will not critically go beyond the established ones, and our prognosis is based on that.
Looking at the picture we can assume that the currencies will get the inertia and this trend will develop in the weeks. But again, this is just a supposition, which by the way, isn't deprived of one hundred percent logical conclusion.
Ok, we will see. I'm just going to take a little time out. I need to get my thoughts together.
You can't trust anyone, not even yourself.
Not believing in yourself is stupid. If you don't believe in yourself, there's no point in doing anything.
It is possible to assume.
Assuming is moving into the "reasonable" field. It is always useful :o).
I have seen such a TS - it predicts the average point of the instrument for the next day (high-low)/2.
But in it the price dynamics is formed by the information coming to the market (macro news, wars, earthquakes, etc.). Accordingly, the confidence interval for the weekly price forecast looks like an expanding funnel and the forecast for Friday's price is absolutely uninformative. Therefore, only the forecast for the next day was really used.