From theory to practice - page 709
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
Genius.
It is too early to open monitoring - so far the results for Variable Gamma process are poor. Somewhere I have an error due to misunderstanding of expectation calculation. Now I'm trying WMA, but I'm afraid it's not the same...
Now look, Dimitri.
Out of all stochastic process models only two tend to "return to the mean" - Ornstein-Uhlenbeck process and Variance gamma process. Both of these processes are described in this book... So it turns out - ALL stochastic process models on the market - to hell! Right?
If so, then this topic should be scrapped, and a new one should be opened - for non-random processes in the market, processes with "memory".
Yes, the topic, in my opinion, does not bother anyone. It even helps.
And as for "genius" - think about it: a bunch of grant-aid institutes puffing away overseas, predicting various economic indicators on a weekly basis. The results are a finger in the sky. Or rather, the best model for their predictions (the best approximator for their various garbage and other econometric nonsense) - tomorrow will be approximately the same as yesterday. This is despite the fact that they usually have maximum input data.
Have we already sent Gunn "to the furnace" as well?
Have we already sent Gunn to the furnace, too?
There's simply no time for it. Although, there's a lot of interesting stuff there - but it's complicated.
If Variance gamma process doesn't work out (and I'm counting on it, despite everything - it really builds the variance channel without any quantiles!), I'll go back to it.
It's kind of hard to do it alone... I think with horror - do I have to be engaged in Forex for 50 years to start earning kopecks? God forbid!
Please post literature on non-markov processes here, until a new thread is started. It doesn't matter what language, even ancient Papuan.
I hope this literature will help find the key to "memory" of the market.
Let's start with Shelepin (see attached archive). + some strange guy's book (but it is interesting to read).
Please post literature on non-markov processes here, until a new thread is started. It doesn't matter what language, even ancient Papuan.
I hope this literature will help find the key to "memory" of the market.
Let's start with Shelepin (see attached archive). + a book by some bizarre uncle (but, read with interest).
Why do you need literature? It is better to think with your head.
The market memory is stable, the price remembers its average condition at any part of the chart. Absolutely at any part of the chart! What other memory can be better than this one?
What kind of memory are you looking for? Explain it to me on your fingers like a physicist to a nerd.
Why do you need literature? It is better to think with your head.
The market's memory is stable, at any part of the chart the price remembers its average state. Absolutely in any part of the chart!!! What other memory can be better than this one?
What kind of memory are you looking for? Explain it to me on your fingers like a physicist to a nerd.
Let me explain.
It's already clear to everyone that a channel strategy around a sliding expectation is the best you can come up with.
Technical problems (expectation calculation, confidence interval limits) are omitted - each random process model has its own requirements.
But ALL models have one crucial drawback - when the price leaves the channel borders, it is unclear where it will go next. The Ornstein-Uhlenbeck model says - back to the mean. But, this is not the case in the market, even though the correlation coefficient, on average, is negative.
So, in my understanding, memory is a measure of data structure. The difference between the current state of the system and the chaotic state. I.e. when, for example, a person remembers where he/she needs to go and purposefully goes there forming a trend - structure of directed movement.
The question is whether the price has moved out of the interval - has the structure been finally formed or not? Is the trend over or has it just started?
That is, all trading strategies in the channel should have an additional parameter responsible for the structure of the data.
What is that parameter?!?
I don't know!
Maybe it's non-entropy... Maybe something else... I have no idea. I haven't found it - and this thread has lost its unrestrained meaning... There's no way without this parameter. Well, there's no way. I'm crying on Schrodinger's cat's shoulder again.
So, how to directly calculate the non-entropy - I do not know.
But, some suspicious guy in his article claims that the analogue of nonentropy is dissymmetry, i.e. the difference between the current distribution and the symmetric one. It follows that the sign of the end of the trend is a strong skewness of the probability distribution.
Here, I included Pearson's asymmetry coefficient into my TS again. Once again I want to see how it works.
Explanation.
It is already clear to everyone that a channel strategy around a moving expectation is the best that can be devised.
Technical problems (expectation calculation, confidence interval bounds) are omitted - each random process model has its own requirements.
But ALL models have one crucial drawback - when the price leaves the channel borders, it is unclear where it will go next. The Ornstein-Uhlenbeck model says - back to the mean. But, this is not the case in the market, even though the correlation coefficient, on average, is negative.
So, in my understanding, memory is a measure of data structure. The difference between the current state of the system and the chaotic state. I.e. when, for example, a person remembers where he/she needs to go and purposefully goes there forming a trend - structure of directed movement.
The question is whether the price has moved out of the interval - has the structure been finally formed or not? Is the trend over or has it just started?
That is, all trading strategies in the channel should have an additional parameter responsible for the structure of the data.
What is that parameter?!?
I don't know!
Maybe it's non-entropy... Maybe something else... I have no idea. I haven't found it - and this thread has lost its unrestrained meaning... There's no way without this parameter. Well, there's no way. I'm crying on Schrodinger's cat's shoulder again...
I see. You have only a vague idea of market memory.
And the price knows where it is going, but that is not memory. More precisely, it is the banks that know in advance (where to go :)) the demand for currency, and since financial flows are, after all, all interconnected, the price rests at the demand or oversupply boundary. These boundaries are quite clear if observed multilaterally.
Many physical laws do not apply to markets and lead many people into a forest of delusion.
I see(. You only have a vague idea of market memory.
And the price knows where it is going, but that is not the memory. More precisely, it is the banks that know in advance (where to go :)) the demand for currency, and since financial flows are, in the end, all interconnected, the price rests at the demand or oversupply boundary. These boundaries are quite clear if observed multilaterally.
Many physical laws do not apply to the markets and lead many people into a forest of delusion.
Maybe the parameter we are looking for is the sellers/buyers ratio? But, how and where do you get this data in real time in your TS?
passing by, decided that Schrodinger's cats do not need to look for a black cat in a dark room, it simply may not be there (C)
окончательно сформирована структура или нет? Закончился тренд или только начался?
a local maximum (minimum) or a global maximum (minimum) will be formed and nothing more, time will show which was the global or local one
well, the trends ... you will draw the line you want, if this line defines something or decides on something