Discussion of high-frequency trading on MT5 - page 34

STP is essentially just a copycat of trading signals. For example, the Signals service is a special case of STP.

A copyer is not necessarily a signal repeat. It can be the opposite action and many other things. But more often than not the signals are exactly copied.

The copier itself is a special case of HFT implementation. Since HFT is always a task to catch a signal.

I wrote above:

The developers have not carried out such an experiment, but we anyway got the graph of dependence of profitability of the "Signal" on the ping:

In the long run it is obvious that the limitation will need to be removed:

Within the Signals service this restriction is only a competitive disadvantage. But within the STP ideology it is a serious disadvantage. Therefore, facilitating the implementation of STP is closing a gap which needed a solution. And such a solution has been implemented many times by third party developers.

So when it comes to

then nothing will really change at all in the realm of broker overlaps, as everything has been working for a long time.

This is generally a good example of developers' foresight. Or rather, sometimes its absence. When traders are constantly talking about the importance of some thing for trading, the developers bend their guru-lines. And when they come across it in practice after several years, they start to move. On the one hand it is natural, but on the other hand it is too arrogant. As a small result of this approach received:

When they throw around the phrase "pipsing is evil", they have little understanding of its depths. In order for an ECN/STP broker to have good order execution, the ECN/STP broker's system must be implemented through the HFT-principles. Because for a broker, execution quality is a huge competitive advantage. And if you even from the screen with a mouse you want to conduct a trade, the result of it will directly depend on the quality of ECN/STP system realization, and more exactly, on sharpening it for HFT. This is something that no one sees. But they start to criticize it when they receive negative slippages and the like.

No one is saying that HFT is a grail. It's just a proper technological implementation of trading signals.

P.S. We are talking about some HFT grail only for the following reasons:

• One:
  Stat. arbitrage carries less risk the closer you get to the noise.
• Two:
  TC by increasing risk can be (roughly and conditionally) divided into the following categories (FOREX):

  1. Arbitrage is the least risky strategy, there is relatively little liquidity. Used by hedge funds and investment banks. Stability is highest.
  2. Noise trading - higher risk, but more liquidity. Hedge funds are rarely used. Stability is significantly lower.
  3. Others (thick-skinned) - highest risk, ocean of liquidity. Stability is lowest.
• Three:
  If you entered at a time when prices didn't change for a second or two, it's definitely not optimal.

What is the point of this post? :)

Misconceptions, Part 2: Statistics is Pseudoscience, or The Chronicle of a Diving Sandwich was an article read. Thought it was interesting. Quote from it:

The article was writtenon 16.05.2008 by:Sceptic Philozoff

Logical conclusion-question: pipsing was not allowed earlier (in 2008), now it is?

Several hundred trades a day and an EA of Lucky type is now in the forefront? I'm asking seriously, without any irony or derision. I wonder if the situation has changed fundamentally in this respect.

Told you before:

These limitations were mitigated by running several terminals to emulate asynchrony.

Today gave my monthly limit slippage statistics for just one symbol (> 15,000 executed limits) + comment.