Future price stack with Time & Sales and transaction display - page 11

New comment
 
olyakish:

It is not quite clear what exactly the deal will be.

Option A

I'm out of the market and I want to buy and I eat somebody's sells with my market order

i end up buying the person who sold either got out of the market or has a sell position

which will be broadcast - mine or the other side of my deal?

Option B

both deals will be broadcast according to the plaza II specification ftp://ftp.micex.com/pub/FORTS/Plaza2/bak/p2gate_ru.pdf

and there will also be information about the part that was reregistered from the big MM limit

approximately here

Received;ExchTime;OrderId;Price;Amount;AmountRest;DealId;DealPrice;OI;Flags

05.12.2014 20:46:13.221;05.12.2014 20:46:13.272;14004005502;0.8268;10;0;991923090;0.8268;46;Fill, Buy, Quote
05.12.2014 20:46:13.221;05.12.2014 20:46:13.272;14004003983;0.8268;10;990;991923090;0.8268;46;Fill, Sell, Quote, EndOfTransaction

I didn't get it either, but they say it will be ok and there is no need to raise panic, and if you don't raise it now, it will be too late later. Which to me, I think the bleakest prospects, can be expected from option A
 
Renat:

It depends on the liquidity provider/exchange the broker uses.

If the provider broadcasts execution prices, the broker cannot prohibit them. Not only that, but full ticks streams will soon be automatically available directly in the MetaTrader 5 strategy tester, which will raise the quality of testing. If the original ticks are available, the price modelling algorithm will not be used, but real ticks will.

Drop MetaTrader 4, the technology has changed!

Thank you, this is a very important update. But many brokers have even one minute history of terrible quality, I am afraid the quality of tick history will be even worse, and then the whole point of such testing will be lost, because it may be even worse than simulated ticks. For example, mt5 terminal should save all ticks locally, and use them for testing. And accordingly, it must load ticks from the broker for those periods when the terminal was "off".

 

To solve the problem globally, we need to make "story quality" a significant point in the brokers' marketing programme.

Unfortunately, they do not see the problem any other way. They can easily spend millions of dollars on advertising, but they do not want to do the tedious manual work of collecting and presenting the quality history with total expenses of 10 000 dollars. Only customer churn and tough claims can shift the focus.

 
Renat:

To solve the problem globally, we need to make "story quality" a significant point in the brokers' marketing programme.

Unfortunately, they do not see the problem any other way. They can easily spend millions of dollars on advertising, but they do not want to do the tedious manual work of collecting and presenting the quality history with total expenses of 10 000 dollars. Only customer churn and tough claims can shift the focus.

And why can't the collection of a quality story be made independent of the broker.
 
sandex:
And why can't the collection of quality history be made independent of the broker.

Because the problem has to be solved globally for the masses, not for the individual.

I am not talking about the need to have absolute consistency of this history with all the settings of a particular symbol/broker. We are making a complete solution - "press Start and everything automatically worked", not "a constructor with crutches and a huge manual for 30 preparatory steps before testing".

 

Well, yes, a lot depends on the broker's attitude to the problem. But until their clients start complaining, it's unlikely that anything will change on their part. I understand that most brokers don't really care about history and keep it purely for show.

A person may, for example, run a test on one broker and see that the quality is 95% and the other broker 90%. In both cases everything seems fine, the percentages are big, and the user does not care. But the test result will be bad. Instead of percentages and green-red picture of quality - you can write about ticks in more detail. For example, "7000 minutes (25%) have no tick data in the selected history interval. The test result will not be correct". When you reach the threshold percentage - mark it in red, or otherwise draw the user's attention that it's bad. Maybe then they will start complaining to the broker about the history.
Imho.

Renat:

"constructor with crutches and a huge manual for 30 preparatory steps before testing".

That's how Tickstory works for generating quality history for MT4 :). But who needs it, goes through all the steps. I think that eventually there will be some custom solutions for MT5 as well, how to achieve the high quality testing with bypassing the history of your broker. Even with crutches, but there is a need for it.
 

We have already made History Center for MT4 - it turned out to be complete rubbish, constantly conflicting with brokers' data. We just listened to traders saying "let us load the data ourselves". We will not make such mistakes again.

Wait for the new release of MT5 with ticks, please. For our part we will also start to stimulate the brokers, as it is an important problem.

 

The benchmark topic is blatantly a bit far-fetched.

Also, you are completely ignorant of the root of the problem of disaggregated tick data:

  1. An in-depth history of the stock markets costs money. All the more so, the ticking. It's just frowned upon by anyone who works in this field.

  2. All stock exchange data is laden with complex non-disclosure prohibitions. The lawyers are just waiting for someone who has money (they don't pay attention to the small stuff, knowing that there is no money to be made there) to make a mistake and start distributing this data. They don't even need a justification, it's enough that you passed by a mile away.


The first person to put out extensive tick data of an exchange for free will ruin the business of selling that data, get a bunch of lawsuits and be driven to bankruptcy. This is a rough model of what will happen. That said, there are a very small number of exchanges that provide their data for free.

 
All this is not bad, it is only the process of trading and deciding how to enter the market that is unclear.
 
papaklass:

My posts related to forex.

I will second the question about tick history for forex. I myself have not seen any broker selling tick history for it (if there is such, please give me a link). However, suppose it exists and MQ releases a new build, where the tick history is available in the tester (at least for the server MQ). It turns out that there is access, just not for all. But who prevents the programmer to start the tester and copy from there all the necessary tick history in any format and distribute anywhere? It will turn out that MQ will threaten the business of selling tick history one way or another. Or am I wrong?
1...456789101112131415161718...20
New comment