Why is the normal distribution not normal? - page 32
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
// minus the expectation to zero out the middle of the bell
NOOOOOOOOOOO!!!
No need to zero in any way, I was joking!!!!!
Exactly without zeroing, divide the two bells, with the expectation shifted. it's better.
It's theoretically possible to calculate. There seems to be a Cauchy kind of thing going on, if I'm not mistaken.
Yeah, that's right.
It occurred to me (about the distributions): if we take a pair of Random generators and divide one by the other (exclude generation of zeros in the "denominator", self-sampling), we get something like the observed one. That is, a pole in the middle and thick tails. With envelope similar to a pair of hyperbolas. And it's understandable - after all, currency pairs are fractions in fact....
Who has everything ready for pictocomposition (c), can you already try it? :)
// I don't claim that they are actually OK. What I don't know, I haven't tried.
// But to start with approximately normal - sum (6-8 pieces) of several usual linear randoms
// minus expectation, to zero out the middle of the bell
Quite plausible. I'll have to check it out.
Mathemat 08.12.2009 21:02
It's theoretically possible to calculate. There seems to be a Cauchy kind of thing going on, if I'm not mistaken.
Yeah, right.
Well, does it look like it?
// Finally, the answer to the "main question of the thread" has been answered!
// Oddly enough ...... :)
Not at all. The price one buys at can be "unfair" :)
Taki, you know, I agree with you, and very much so :o)
Плотность распределения Коши - это функция типа 1/(1+х^2). Но распределение returns все же имеет не настолько толстые хвосты.
I think this is how it should be. In other words, the plan is to generate these things and compare them thoroughly with what we actually get from the quotes. The difference is cashed. If there is no difference, we leave Forex. :)
I think that should be the case.
So, the plan is the following: we generate these compilations and thoroughly compare them with what we really receive from the quotes. The difference is cashed. If there is no difference, we leave Forex. :)
I'll comment on the highlighted. // The rest is a bit of an almost joke. :)
Pairs are primarily traded. This trading is rather chaotic and generates a tendency for retuns to be "normal".
Arbitrage processes, on the other hand, trade specifically currency indices. Which encourages the distribution to pull in the direction of the koshi.
But the arbitrage process is not omnipotent and has all sorts of limitations : unprofitable spread channel, time delays of various nature and probably something else (laziness and stupidity of traders for example).
Therefore, the observed hybrid is the result.
Taki, you know, I agree with you, and very much so :o)
)))
I think this is how it should be. In other words, the plan is to generate these things and compare them thoroughly with what we actually get from the quotes. The difference is cashed. If there is no difference, we leave Forex. :)
Your idea is very good (I mean the idea). But I don't understand the implementation... I'm tired or something. Tomorrow I'll reread it and try to comment on it.