FR H-Volatility - page 7
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
Of course, I don't mind writing. I don't always manage to Skype, because I get home late and don't want to wake up my family. But I can do it on ICQ, it's silent.
ICQ, so ICQ 213-878-996.
About the same as Peters did in his book http://bigforex.biz/load/8-1-0-136. You can also read http://bigforex.biz/load/8-1-0-137, it's his first book. In my opinion, what you call perturbations fits quite well into his model, as he does not single out these perturbations in any way, but simply considers the return process without cutting out the trend sections from it.
I know the task is very difficult, but the goal is very tempting. By the way, Peters is not the only one. There's also Shiryaev, who also studies financial series statistics.
I read Peters about two years ago. It's all right, it's an interesting book, it broadens my horizons. However, its usefulness, both in terms of understanding the nature of the market, and in terms of practical methods of working in it, alas, is very limited. Maybe it is because "his model" does not even try to look inside, i.e. does not rise above phenomenology. And Shiryaev, as a mathematician should, does not go inside the phenomenon, but investigates it "in its entirety", as it is. That is why (with all my respect to Shiryaev, a recognized coryphaeus of Soviet mathematics) he does not even aim at uncovering the laws of the market.
I was talking about the physicist's approach, and expressed my own opinion, without imposing it on anyone. It was a physicist Einstein who calculated the run of a Brownian particle, and a geophysicist Hirst who calculated the spread of filling of a reservoir. And the distribution of energies of ideal gas molecules is called the Maxwell distribution. So I'm thinking along the same lines. Mathematicians, of course, solve different problems and their approaches are different. And it is good, as they are two different sides of one fundamental process.
As for statisticians, you are absolutely right. In the Stochastic Resonance thread, when I posted my work, I asked a question about the FR being investigated. There was no answer to it. Yes and there were only three attempts. And it turns out it is a special case of a known and researched function called Gamma distribution. I came across it by chance, while reading a book on Bayesian statistics.
This is where it gets interesting, Yurixx. It seems to me that with our level of access to the market we are simply doomed to a phenomenological description at most. Roughly speaking, classical thermodynamics rather than statistical thermodynamics. Doesn't the classical one work well? Even if we don't understand entropy or temperature very well within its framework, it still works, and very well at that.
Ultimately, the true reasons for what such hidden internal reasons for TC work will still remain behind seven locks for us. The marketplace is a black box. The key is to learn to recognise what that box will produce in the next moment (or the week after) and to make sure that our knowledge extraction system is working steadily.
Yurixx
...how to make a CB generator that works according to a given distribution ?...
You need to have an analytical expression of the distribution function F(x) and there is an inverse function F^-1(x) then it's simple. Or at least there is a procedure in math software that calculates this function.
Mathemat
...
P.S. Wouldn't it be nice to build bars with equal number of ticks rather than with equal astronomical time in them ...
...
I'd really like to see it too, if you come across it, be sure to knock me.
The funny thing is that the Close process of such "equivolume" bars may well turn out to be something like Wiener's, as the tick returns FR has almost perfectly two sharp peaks at +-1 (but that's for the EuR). It turns out something very similar to what Bachelier dreamt of in 1900 (and turned out to be right - with a time correction breaking the whole FR). But this hypothesis needs to be tested.
In principle, there is an article on this subject: 'The principle of substitution of time in intraday trading'.
P.S. Can you imagine what it means? In a market presented in this form there are no disasters, i.e. no fat tails - as everything is Gaussian (with very rare exceptions, boiling down to less than 1% of all ticks greater than 1 modulo). Yurixx, it turns out that your joke about the vinarity of the process isn't that far from the truth. All you have to do is change your glasses...
P.P.S. Rosh, do you still remember the idea of converting a real FR to a Gaussian? And here you don't even need to do that...
P.S. It would be nice to build bars with an equal number of ticks rather than equal astronomical time in them...
Will the end justify the means? ;)
P.P.S. Rosh, do you still remember the idea of converting a real FR into a Gaussian? And here you don't even have to do that...
I once analyzed those normally distributed values (on which it was so easy to make money without any analysis) given to me by Excel about a month/two ago for Z-count. And found out that this generation was not unconditionally normal, it clearly had dependencies between increments (returns). So, you can continue studying, no solution has been found yet :)
Of top, but don't know where to write. If I am not mistaken the Guinness Book of Records has a record of 1200% p.a. Larry Williams http://web-investor.academ.org/index.php?action=articles&id=71
Looks like they made it. If I were the organizers I would invite observers from that organization and if they confirm the record IHMO good publicity at least.
komposter
The end justifies the means, but only if the aim is noble. I'd like to have 1 eye on this stream. And with my crooked hands it won't happen soon :-)
Rosh, in principle it is possible to mix these data (Gaussian process). Then the dependence should kind of disappear.
2 komposter: God only knows if it's worth the effort. I think it's not hard to make an indicator. But it seems that on such charts all sorts of wizardry like Fibs and calipers/resistants will stop working. But it is quite possible that the simplest critters like wipers, RSI, stochastics and others will start working better.
The end justifies the means, but only if the end is noble. I would like to have one eye on this stream. And with my crooked hands it won't happen soon :-)
I'm waiting for a detailed description on Skype.