FORTS: Strategies and how to implement them - page 14

 
anonymous:

Don't show off.

OK.

:)

Go back to your quickie with Prival. get your kinks done there.

 

There, another sign of the degradation of the forum.

They used to be able to tell the clowns from the decent ones. And now Prival and Anonymous are scorned and Sulton is regarded as a respectable forum member )))

 

http://www.russian-trader.com/forums/threads/3444-%D2-%CF%F0%E0%E2%E4%FE%EA-%CC%E0%F2%E5%EC%E0%F2%E8%F7%E5%F1%EA%EE%E5-%EE%E1%EE%F1%ED%EE%E2%E0%ED%E8%E5-%EF%E0%F0%ED%EE%E3%EE-%F2%F0%E5%E9%E4%E8%ED%E3%E0

There's one about logorrhythms ))) This is a good article for those who are interested in arbitrage and pair trading. I advise those who are interested in arbitrage and pair trading to read this article.

Т.Правдюк: Математическое обоснование парного трейдинга
Т.Правдюк: Математическое обоснование парного трейдинга
  • 2010.09.30
  • mehanizator
  • www.russian-trader.com
В парной торговле трейдер делает предположение о синхронности движения высококоррелированных активов. Акции компании одной отрасли должны, согласно этой теории, реагировать одинаково на внешний фон, если он не затрагивает непосредственно только одну из них. Динамика цен повторяет друг друга, а любое отклонение от привычного паритета должно быть...
 
iron_056:

There's one about logorrhythms ))) I have some tips for those who are interested in arbitrage and pair trading. You can find a lot of interesting and useful articles of the author on this web-site. I advise those who are interested in arbitrage and pair trading to read them.

Note!

- when searching for cointegrated pairs it is necessary to work with logarithms of prices, because otherwise the quadratic trend will introduce a strong distortion;

- When identifying pairwise correlations it is desirable to work with increments in the logarithms of prices, because the presence of a general trend in the stock market can unreasonably overestimate the degree of correlation;

 
anonymous:

2. Look carefully at the prices at which the C-4 rate is calculated on page 12 - it is "high". 12 - it is 'high'.

You are picking on it. Yes, the prices are out of sync, but the accuracy is still high. I can calculate "in sync" on minutes, the result won't change much.
 
TheXpert:

There, another sign of the degradation of the forum.

They used to be able to tell the clowns from the decent ones. And now Prival and Anonymous are scorned and Sulton is regarded as a respectable forum member ))))

Andrew, please do not confuse things. I don't understand why you should always try to solve the problem of positioning of the market, but if you want to succeed, you've got to understand that the market is more profitable. All this movement in the branch is an enticement of clients, you will soon see for yourself.
 
TheXpert:

There, another sign of the degradation of the forum.

They used to be able to tell the clowns from the decent ones. And now Prival and Anonymous are scorned and Sulton is regarded as a respectable forum member ))))

No one is badmouthing anyone. It's just that the complexity of the wording is no indicator of the degree of meaningfulness.
 
joo:
Andrey, please do not confuse everything. At least the claim that it's impossible to test tickwise strategy in a specialized trading language, and at the same time, the example of hand-made shuffle-like creations must cause rejection, to put it mildly... All this movement in the branch is an enticement of clients, you will soon see it for yourself.

The rejection is not a specialised language for algotrading. It is the format of the story you can work with. Are you going to test ticks in MT?

Do not make me laugh, it's not there, there are minutes, built on prices last...

I, unlike you, can test it on the history of the market. You have to wait for MQL for at least another 10 years.

 
anonymous:

A vegetable garden needs to be planted so that the resulting process is interpreted not as "price difference, ololo", but as an interest rate.

No argument, models are needed. And your comments in this area are some of the best on the forum (no sarcasm). But sometimes when building a model you lose touch with reality. You need to trade the market, not the model that supposedly describes it.
 
Prival-2:

I, unlike you, can test it on the history of the market. You have at least another 10 years to wait for the MQLs.

I can also test any strategy on the market history. Believe me, there are no more fish there than in any other trading place.