Put in a good word about the occasional wanderer... - page 4
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
движение цены совершенно не предсказуемо. мы имеем дело не с математикой, а с психологией, и тут никакие формулы не помогут
Psychology (as a set of rules of human behaviour) is the easiest to formalize,
The hardest thing to formalize is madness (it's like a monkey with a grenade, you never know when or where he's going to throw it :o)
In any time interval the SB will have a normal distribution; from 1015 to 2256 or from 1305 to 5321. In general, any segment of variable length will give a normal distribution.
I've written that myself ten times. But it is of fixed length, not variable
What distribution then do you think the SB has, isn't it non-stationary? Move away from these increments, look at the process from a different angle. If you see a clearly bounded bell, it does not mean that the process forming it is stationary.
The fact that SB is unsteady is a fact. I gave a link where this was described. SB is an unsteady I(1) process.
Psychology (as a set of rules of human behaviour) is the easiest to formalize,
The hardest thing to formalize is madness ( it's like a monkey with a grenade, you never know when and where he will throw it :o)
The psychology of one person or group of people under specific circumstances can be predicted. There are billions of people with all sorts of circumstancesSecondly, this is exactly why statistics can be applied.
я это уже сам раз 10 написал. Но именно фиксированной длины, а не переменной
Once again, no, exactly variable length. Starting from any point in the SB at infinity, the distribution will be normal.
Answer the question: "What is the distribution of the SB process?
psychology of one person or a group of people under specific circumstances can be predicted. There are billions of people with all sorts of circumstances
It's exactly the opposite. It is impossible to predict the behaviour of one particular individual. On the aggregate level, however, the behaviour of a crowd of many individuals is much easier to predict. Advertising, electoral technology, marketing, etc. are built on this.
Всё с точностью до наоборот. Невозможно предсказать поведение одного конкретного индивидуума. Зато на агрегированном уровне поведение толпы из множества индивидуумов предсказывается гораздо проще. На этом построены реклама, выборные технологии, маркетинг и пр.
That is where we stand, so the essence of trading is to recognize the current behavioural pattern and
Make a trading decision based on the knowledge about its evolution,
The second task is to statistically find the best decision points of similar models.
To make it easier (not to identify a specific model but a class at once).
Answer the question: "What is the distribution of the SB process?
In principle this is where https://www.mql5.com/go?link=http://hometask.boom.ru/economics/econometrica/5.html describes it all quite well.
The conclusion will change if we consider the process from a certain point in time, e.g. from t = 1. Suppose that Y0 is a deterministic quantity. In this case the process AR(1) will not be stationary by the above definition. The variance of Y and the autocovariance will depend on t:
var(Y t) = s , cov (Y t,Y t-t) = ct t .
However, over time such a process (as long as êr ê< 1) gets increasingly close to stationary. It can be called asymptotically stationary.
P.S. There is also the formula SB Y t = m +r Y t-1 + e t, t = (-¥,...,0,1,...+¥) (assuming that e t ~ IID(0,se2) are independent equally distributed random variables with zero expectation and variance se2).
P.S. there is still a sense to talk about increments, because the author formulated the problem exactly through increments
В принципе вот здесь https://www.mql5.com/go?link=http://hometask.boom.ru/economics/econometrica/5.html все достаточно хорошо описано.
Вывод изменится, если рассмотреть процесс с определенного момента времени, например, с t = 1. Предположим, что Y 0 — детерминированная величина. В этом случае процесс AR(1) не будет стационарный по данному выше определению. Дисперсия Y и автоковариации будут зависеть от t:
var(Y t) = s , cov (Y t,Y t–t) = c t t.
Однако со временем такой процесс (если только êr ê< 1) все больше приближается к стационарному. Его можно назвать асимптотически стационарным.
P.S. смысл есть все же говорить о приращениях, т.к. автор сформулировал задачу именно через приращения
Now that's called forgery. The question was about random rambling and you inadvertently switched to a mean-reverting process, which, as they say in Odessa, is two big differences.