a trading strategy based on Elliott Wave Theory - page 264

New comment
 
I don't believe that common strategies can be profitable

hmm. not true. (why then do wave traders exist?)

in fact, you need at least the total MO of all bank traders using it to be greater than zero.
in fact the total iOp of all bank traders using it should be at least zero. and you can count on your fingers the number of people making profit with tactica adversa: the system is not simple (although it uses elementary laws of physics), so 99% of those who want to try it have no patience to grasp it completely...

most people are still in the first stage: "I wish I had a computer sooner, then I'll be a millionaire in six months"... and sometimes for decades... ...everyone does, but not everyone does. (eh... it was a fun time...)
 
The first thing Yandex came up with:<br / translate="no">http://lib.irismedia.org/sait/forex-kiev/multi_fr.html

No, that's not it.
Einstein, by the way, most of his life was engaged in the unified field theory. And it seems that he also could not offer anything complete :)

good point... convincing... perhaps the (Mis)behaviour of markets is worth reading after all...

I should probably clarify: Einstein really did spend most of his life unsuccessfully pursuing a unified field theory.
I do not believe that common strategies can make a profit

hmm. wrong. (Then why do the wave-makers exist?).

Wouldn't it be more correct to ask: At whose expense do they exist? :)
in fact, you need at least the total MO of all bank traders using it to be greater than zero.

Only it is necessary to take into account the weight, i.e. the capital, which a trader operates with. However, I suspect that professionals do not use Tactica Adversa, i.e. it does not affect the market. Maybe one day they will get around to it.
 
Tovaroved:
the first thing that came up on Yandex:
http://lib.irismedia.org/sait/forex-kiev/multi_fr.ht

On reflection, I interpreted it as "Read the manual at last" :)
After a short search I found an equivalent 1/f noise diagram:
 
thank you... yes, it's a sensible model. the only question for me is how to work with it.

and the link to the Mandelbrot article... he writes there:
Sight tells us that these three graphs are unrealistically simple. Let us now open the sources. Figure 1 illustrates the price fluctuations according to the model introduced in 1900 by the French mathematician Louis Bachelier. The price changes follow a "random walk", which corresponds to a bell curve, and illustrates the model underlying modern portfolio theory. Graphs 2 and 3 are partial refinements of Bachelier's work: one model I proposed in 1963 (based on steady-state random processes) and another I published in 1965 (based on fractional Brownian motion). These variants, however, are always inadequate, except for some special market states.

In the original:
The eye tells us that these three diagrams are unrealistically simple. Let us now reveal the sources. Chart 1 illustrates price fluctuations in a model introduced in 1900 by French mathematician Louis Bachelier. The changes in prices follow a "random walk" that conforms to the bell curve and illustrates the model that underlies modern portfolio theory. Charts 2 and 3 are partial improvements on Bachelier's work: a model I proposed in 1963 (based on Levy stable random processes) and one I published in 1965 (based on fractional Brownian motion). These revisions, however, are inadequate, except under certain special market conditions.


fractional Brownian motion, is it not 1/f? if not, sorry, wrong...


P.S. but i would read einstein anyway. better an erroneous theory with new ideas than empty incoherent formulas repeated a hundred times by people for whom they mean nothing but letters...
 
If we open a LOT of instruments at Long, then after a certain time, with a numerical equality of price increments to the 'plus' and to the 'minus' due to the difference in the absolute average increments of rising and falling stocks, we will come out in a stable plus!


to Rosh 05.04.07 11:07
Neutron, in fact the situation is the same with currencies...


Yes indeed the same effect is observed on currency instruments - the volatility is directly proportional to the value of the asset: sigma0=a*Bid, where sigma0 is the daily volatility, a is the proportionality factor.
It is not difficult to evaluate the profitability of the strategy based on the above effect. The difference in absolute values of increments of falling and rising instruments will give our daily return in points and it must be compared with the difference in swaps of short and long positions that averages 2-3 points.
So, at the end of the trading session the absolute values of the long positions increments are dLong=a*(Bid+sigma0), short - dShort=a*(Bid-sigma0). Profit per day: S=dLong-dShort=2a*sigma0.
For currency pairs, the proportionality factor is 1%, sigma0 is 100 points/day, S=2*0.01*100=2 points/day, i.e. if the difference in swaps is 1-2 points/day, we most likely cannot make profit!
About the same situation in CFDs and Futures - a=1%, sigma0=30-100 pips/day.
 
Thank you... yes, quite a sensible model. the only question for me is: how to work with it? <br / translate="no">
If you mean how to describe the processes in it, it's chain theory. If you mean how to correctly predict the market based on it, probably no one knows it :)
fractional Brownian motion, is it not 1/f? if not, sorry, my mistake...
As far as I understood it, it's a special case of 1/f (Gaussian fractional noise). But in general the definition game is a kind of scholasticism, and to understand what Mandelbrot was doing in 1965 you have to read his work. Indirectly, we can conclude that he was still dealing with 1/f noise because the definition mentioned in the quotation refers to the 1982 paper.

P.S. but i would still read einstein. better an erroneous theory, but with new ideas, than empty incoherent formulas repeated a hundred times by people for whom they mean nothing but letters...

Well, in short, yes. Although there is a danger, under the magic of the great ones, of missing the fork that avoids the stalemate.
Using FFT by klot I quickly took a spectrum of a random EURUSD plot and saw the most natural 1/f
 
1/f is flicker noise. http://investo.ru/forum/viewtopic.php?p=187204#187204
http://www.investo.ru/forum/viewtopic.php?t=69871&postdays=0&postorder=asc&start=800

Yes, looks interesting, have to read it. Thanks.

Added:
On first reading liked it. Only the assumptions that look indisputable are used, then the technically correct actions. I really just want to poke around in things that look vague :). Although the thought "could not implement something similar, the more so described in sufficient detail" naturally arose.
 
Hello, everyone! I've missed a lot of interesting stuff while I've been away on business. Got a bit flushed for my model. Lots of maths, but no analogy with electrical circuits, strings, mechanisms, and other directions. Experience is a useful thing, but I think that way is wrong in this case, although the history of physics may remind you otherwise. Once I chose the following saying as a motto for my model (sorry for pathos :o):

<br / translate="no"> "Evolution... the basic condition to which all theories, hypotheses, systems must henceforth obey and satisfy, if they are to be reasonable and true"...

Pierre Teilhard de Chardin.

But this is my opinion. Shame, catastrophically lacking time to finalize my model, but nothing, somewhere ahead of holiday.

The filter published by Sergey, reminded me of the collected statistics on trends bounded by local extrema with a smoothed filter line. Found a lot of interesting stuff then. Decided to share such useless statistics. The algorithm is very simple:
(1) filter parameters are iterated
(2) performs filtering
(3) local extremum counts are found (minima and maxima are respectively alternated)
(4) Successively selecting trends (signals) bounded by these intervals
(5) Necessary calculations are performed for each trend (channel), depending on the task

The statistics is gathered for all series (H+L)/2 on the clock and over the entire history for a whole bunch of criteria. I am showing you some useless results for EURUSD on the history of 5.7 years.

Frequency of trend occurrence. The data is more or less in agreement with the studies published on the spider


This is the distribution of mathematical expectation according to the trend lengths


Mathematical expectation normalized to the trend length. I forgot to add. No such pattern is observed for the standard deviations.


The same, but in double logarithmic coordinates


Trends energy distribution as a function of length (in terms of DSP)


It may or may not be useful for somebody. :о)
 
Blushing a bit for my model. Lots of maths, but no analogy with electrical circuits, strings, mechanisms, and other directions.

Don't feel bad, if there's even one equation in the model, there's sure to be an analogue, and most likely more than one :)
"Evolution ... the basic condition to which all theories, hypotheses, systems must henceforth obey and satisfy, if they are to be reasonable and true"...

If we divide the mentioned objects into statistical, phenomenological and microscopic ones, the first ones, in my opinion, are the least designed for evolution, since they are entirely based on the past. The latter are probably the most. My last posts are just that... dreams of such a model :)
I also spent a lot of time collecting similar statistics, but then I came to the conclusion that their use in building a strategy is essentially a fitting of history, even if it's more correct than using parameters in the tester. However, this does not mean that it is useless at all.
1...257258259260261262263264265266267268269270271...309
New comment