Discussion of high-frequency trading on MT5 - page 44

 

So are there any people here who are working on the HFT theme?

I suggest that you post pictures of developments.

Files:
25.JPG  256 kb
26.JPG  246 kb
 

ProstoTak:

 Вопрос к hrenfx

Как Вы отличаете "фантомные объемы" (заявки) от реальных объемов (которые проторговываются)?

The result is indistinguishable. The only question is what the mixture is like and whether it can be filtered out.

papaklass:

When trying to download the FDK

you will see a window with username and password. If you don't have them, you should register

Here you register and then enter these data when downloading.

But, today I decided to download data on ticks and I failed. It says there are problems in the network. May be because today is a day off. I downloaded data one week ago without any problem. I'll try again on Monday.

This window is just registration on the site, I registered immediately and logically used this data and then the data from the demo account, neither works.

Here is an example with the registration data on the broker's website:

Adress: FXOpen-ECN Demo Server - FXOpen Investments Inc.

Username: a.b

Password: 1234567890

 
ProstoTak:

So are there any people here who are working on the HFT theme?

I suggest that you post pictures of developments.

Can you describe in a few words what is so amazing in the pictures, or just worth attention?

What kind of development, algorithm, its part, what you can see in the picture, so as not to fantasize unnecessarily?

 
Alex_Bondar:

Can you describe, in a nutshell, what's so amazing in the pictures, or just worth seeing?

What is the development, the algorithm, its part, what can you see in the picture, so as not to fantasise too much?

Let's start with just pictures? And then there will be a topic for discussion.

At least between people who are really involved in this topic and not just passers-by.

 
ProstoTak:

Let's start by just posting pictures, shall we? And then the topic of communication will be mutual.

At least between people who are really involved in this topic and not just passers-by.

Let's not do that. I asked you to comment on the pictures, this is not a photo galleryjust to post pictures.

If you take into account the design, your "picture" does not hold out any criticism, it's a total P, the question is how interesting the data is displayed.

If it's just a price chart with orders and there's nothing unique about it, then hardly anyone will even bother to comment on it. Either there will be ridicule.

 
ProstoTak:
I would like to find (read for general development) articles, research papers and books with mathematical bias on the subject of currency trading in the interbank market using HFT algorithms.
Hello I wonder if you read this link there just studies HFT algorithms if you read it carefully (there are many links on the same resource to various studies without leaving it) and pictures above the roof with a detailed description
Nanex ~ The Rise of the HFT Machines
  • www.nanex.net
It's not high frequency trading (HFT) that concerns us. It's high frequency quoting, and it should concern everyone. The two images below tell the story. The chart on the left shows the growth of high frequency quoting. The chart on the right shows the (lack of) growth of high frequency trading. Quote data is from CQS, trade data is from CTA...
 

server:
Здравствуйте мне интересно вы читали по этой ссылке  там как раз исследования HFT алгоритмов ,если там внимательней все читать (ссылок очень много не выходя с этого ресурса) да и картинок выше крыши с подробным описанием 

Thank you - I'll look into it today. P.S. I won't be writing in the pub anymore.

Good luck and fruitful work everyone.

 
ProstoTak:

Thank you - I'll look into it today. P.S. I won't be writing in the pub anymore.

Good luck to all and productive work.

Yet another comrade leaves us, strange but find it where the same hurt to harbor in themselves unfortunately - and so everything started well (I understand that everyone has a different sense of dignity but it should not affect by definition on our joint research questions that interest us. )

P.S Come back after studying the material interested in your opinion, or discussion of some direction of this trade.

 

Somehow a substantive discussion does not work.

It's pretty clear why.

++++++

Suppose a person decided to do HFT or just build a machine that gives 99% result.

For example, he opens a standard Level 2 or writes a program that gets a quote flow from some source.

Then where to begin?

It is clear that all answers are in this stream, what is the first thing that comes to mind?

To find some basic parameters that are on the surface.

What are these parameters?

The volume of the position, is it important? Yes, it's important.

Gaps between positions, can they give something in the future? Yes, absolutely.

The lifetime of positions, can something be squeezed out? Absolutely

Can we hypothetically calculate the trades of positions? We should learn to calculate the chain of changes in the position volume. It may give something, yes it may.

And there are a lot of such questions that a researcher asks. Using logical reasoning, we can start by finding a dozen of parameters that will give rise to new parameters to build some models from.

The most elementary thing, for example, that can bring immediate results, is distribution of total volumes (bid and ask).

Often the market is either tense (overbought or oversold) and the breakdown occurs at important fundamental events, such as the interest rate announcement. And this works for any currency and cross (if such a phenomenon occurs). After the news appearance, not always, but there is an elementary situation, the price freezes (does not fluctuate much) for a certain short period of time, with a clear (two or three times) volume distribution in a certain direction. So, you can safely enter on the side in which there is a larger volume and leave the position for at least a day or two. The result is 100% without any buts.

This is the most basic and trivial inefficiency (you can monitor symbols, you can make 100% entry 4 or 5 times a month).

There are many such inefficiencies if you are serious about it.

If there is a grail, it is only in Level 2. The market cannot be based solely on the best bid and ask prices to construct anything worthwhile, let alone candlestick analysis invented by someone (charlatanisma that has captured the minds of millions).

For example in Level 2 you can predict 90-99% of market movement by 0.5-1.5 points at the very beginning and then work on improvements.

Why the discussion does not work? The themes are new and there are no common indicators or any common ones in Code Base that cannot be used for Level 2, they need their own mathematics, long observations, experiments, verification of theories and models and the main thing is a lot of free time.

Here it is not quite correct to demand that someone just starts throwing out algorithms, reasoning about them. Maybe the person took months to get to some important concepts, and certainly will not be put it on the public site, as the return will not be anything, as practice shows.

I do not know how to revive the topic.

So there is nothing to do in the public domain, maybe there will be people who will begin to publish their thinking, worthwhile problems and so on. This will be informative, they'll want to participate in the discussion, put their ideas out there, so ......

Good luck to everyone.

 
ProstoTak:

....If there is a grail, it is only in Level 2...

Apparently the glass is fake. I will still do a series of tests, but so far only noise. Therefore, candlesticks, in spite of their alchemy, are more effective, because many people analyse with candlesticks and create attractors thereby.