From theory to practice - page 174
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
СанСаныч Фоменко:
And thousands and thousands of people have been doing it for 30 or 40 years.
And judging by economic forecasting, they will be doing it for another 100 years without the same success.)
I'm going to express a rather specious idea.
If we achieve a perfect coincidence of practical values of increment distribution (by selection of readout time intervals or just by reading each tick) with formula (14) from post #1728 - then we can make predictions based on extrapolation.
And they just have to match and they do! I just haven't gotten close to it yet and don't see it yet.
Here are the charts for AUDCAD over the past 2 weeks with a sample size of 16900 ticks for exponential readout time
Yes, everything seems fine and good, but something is bothering me... Let me explain what.
What's on the top, what's on the bottom?
And what bothers you is that you have successfully (maybe even optimally) filtered out the trend component and will periodically get kicked in the chin by the trend))
What's on the top one, what's on the bottom one?
And what bothers you is that you have successfully (perhaps even optimally) filtered out the trend component and will periodically get kicked in the whiskers by the trend))
I already get it from time to time :))))))))))) That's why I'm searching so earnestly for Hirst's analogue and seem to have found it in the form of non-entropy. But, still need to work in that direction, of course...
And, judging by the results of economic forecasting, for another 100 years they will be doing it without the same success)).
Every sensible thing has its end, and only bullshit can be done endlessly :))) These undoubtedly worthy people put the cart before the horse for some reason. And for forty years they cannot understand why it does not go.
I already get it from time to time :))))))))))) That's why I'm searching so earnestly for Hirst's analogue and I seem to have found it in the form of non-entropy. But, there is still work to be done in this direction, of course...
Every sensible thing has its end, and only bullshit can be done endlessly :))) These undoubtedly worthy people put the cart before the horse for some reason. For forty years they cannot understand why it is not moving.
Only educated people do rubbish, but they cannot do rubbish by definition.
So you are convinced that there is no superposition of different processes and a wave function can be used? IMHO, it can only be done near levels.
I am absolutely sure of one thing: that I am solving the problem correctly. But, being exhausted by work during the day and by my relatives' questions in the evening, such as "When will the money come from Forex? You promised!!!", already fussed and somewhere I miss something and do not work out, of course.
Remember I said that the incremental distribution we see is the product of 2 probability density functions?
I've finally split them up and seen them. Look! This is for the AUDCHF pair.
Blue is the real increment distribution.
Green and red - these are the density functions, whose product we see on the blue graph.
What is it all for, you may ask?
The answer is to calculate the sample volume (the most profitable timeframe).
For example, the calculation for this AUDCHF pair
The red and green charts have crossover points = +-10pips, which corresponds to the 99% confidence level of our real blue chart.
The calculation yields 10,000 ticks collected at exponential time intervals with a starting point of 2 seconds.
These 10,000 ticks are collected in approximately 10,000*2.57 = 25700 sec, which corresponds to about 8 hours.
It turns out that it is more convenient to use timeframes H4 and higher for trading.
How do people manage to earn something on M1? - It's a mystery...
I will try to answer this riddle myself - why some people successfully make money on M1-M30.
If you paid attention, my average frequency of guaranteed tick reception is 1 tick in 2.57 seconds.
Thus, if the Man with capital letter manages to be guaranteed to receive 1 tick in 257ms, then this "bell" of increments is collected 10 times faster from him.
10000x257ms = 2570 sec = approximately 40 min.
I.e. A GUARANTEED tick acceptance of approximately 1 tick in 220ms is possible to build a TS on M30.
However, there are probably few people who can guarantee such reception. There are time gaps between ticks reception and in 30 minutes in one case the necessary 2000 ticks are received, and in the other case only 1 tick is received, and no one even thinks to enter "pseudo-states". Hence the most disgraceful losses of deposits, especially when robots do it.