Put in a good word about the occasional wanderer... - page 5

[Sceptic Philozoff]

FOXXXi >>:

Ответь всё же на вопрос : "Какое распределение у процесса СБ?"Только оставь разность в покое.

Can I answer that, FOXXXi? The lognormal is at the quoted one. This theoretical model is inherent in the option valuation formula. timbo can confirm this as he knows the Black-Scholes formula.

The lognormality distinction has been talked about for a long time - but it is the simplest model.

With SB, I don't know.

[Avals]

timbo писал(а) >>

Now that's called forgery. The question was about random rambling and you inadvertently switched to the mean-reverting process, which, as they say in Odessa, is two big differences.

From the same place:

A first-order autoregressive process at r = 1 is called a random walk. If m = 0 , then it is a random walk in the proper sense, while when m ¹ 0 it is a random walk with drift.

Z.I. I wasn't going to do any faking. But if the definitions differ from yours, then provide links to them

[[Deleted]]

Avals >>:

P.S. там и есть формула СБ Y t = m + r Y t–1 + e t, t = (–¥,...,0,1,...+¥) (предполагаем, что e t ~ IID(0,se²) — независимые одинаково распределенные случайные величины с нулевым мат. ожиданием и дисперсией se²).

P.S. смысл есть все же говорить о приращениях, т.к. автор сформулировал задачу именно через приращения

We look in the book - we see a figure.
This is NOT random wandering, unless p equals 1. The author stipulates right away that it is less than 1. That is, it is a mean-reverting process, not a SB.
It only makes sense to talk about what you know, and if you don't know, keep quiet or ask. Ignorance is not a sin, the sin is militant ignorance.

[[Deleted]]

Mathemat >>:

Можно я отвечу, FOXXXi? Логнормальное. Такая теоретическая модель заложена в формуле оценки стоимости опциона.

Why lognormal? We are not talking (yet) about prices, i.e. our SB could easily go into negative territory. Therefore, it is simply normal.

[Avals]

timbo писал(а) >>

We look in the book - we see a figure.
This is NOT random wandering, unless p equals 1. The author stipulates right away that it is less than 1. That is, it is a mean-reverting process, not a SB.
It only makes sense to talk about what you know, and if you don't know, keep quiet or ask. Ignorance is not a sin, the sin is militant ignorance.

Yes at p=1 SB. And what does it imply that the process is stationary? I wrote that it is non-stationary. FOXXXi asked for a definition of SB there. What's the problem? :)

The first differences DY t of an autoregressive first-order process with r =1 are simply errors e t, i.e. the first differences are stationary. A nonstationary process whose first differences are stationary is called an integrated first-order process and is denoted by I(1). A stationary process is denoted by I(0). If k-e differences of a random process are stationary, then it is called a k-th order integrated process and is denoted I(k).

I wrote the same thing from the second page several times

[techno]

timbo писал(а) >>

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.

No, advertising deals with a completely different topic, and political technologies again work for a limited space. The market is a homogeneous substance at the same time, but it reflects millions of different probabilities of people's choices and behaviour. So it is impossible to predict anything, there is only a fraction of the probability of success, but not 100%.

[Mykola Demko]

Techno >>:
...Так что предсказать что либо невозможно, есть лишь доля вероятности на успех, но 100% нет.

I completely agree with the last sentence, it's a possibility and we're trading.

[techno]

Urain писал(а) >>

I completely agree with the last sentence, it's a probability and trade.

I'm not going to trade with you, but the thing is, the probability is usually 50/50, in most EAs you can swap buy to sell and it won't change much, what matters is stops, trawl, stall stops, etc.

[Sceptic Philozoff]

timbo >>:

Почему логнормальное? Мы же не говорим (пока) о ценах, т.е. наше СБ легко может уйти в зону отрицательных величин. Потому просто нормальное.

OK, fine, let it be normal if the values can be negative.

So what good is that going to do? At first rough approximation, it's still an I(1) process.

P.S. By the way, lognormal has what tails - thick?

Yeah, I see, in first approximation it's something like a degree with a slowly increasing modulo negative exponent.

[Петр]

The crowd trades, and the crowd is a much dumber entity than the individual, and its actions are predictable. It is all clear. And what we see in the form of trends and trendics, is most likely, the crowd's reaction.
BUT!!!
The thing is, the crowd is being controlled! The actions of these "effective managers" are much more difficult to calculate. If they can be miscalculated at all. Here "effective managers" are not specific individuals, but a category of factors (where specific individuals are not excluded, of course.))
Hence the unsteadiness.
Such is the case.