Is it better not to trade at 11 p.m.? - page 4

 
Alexey Busygin:
I still don't get it, how do you calculate?
Read the thread first and you'll see.
 
Server Muradasilov:
Reread the thread first, and you'll understand

In the beginning, everything looks legitimate and civil, but by page 4, it's clear that something isn't right somewhere. And it smells like hacking.

Especially, this phrase, "When started, the guy channeler down the adviser (or crashed by mistake, or UPU died), but the owner of TC not even tempted to monitor this situation.

 
Alexey Busygin:

In the beginning, everything looks legitimate and civil, but by page 4, it's clear that something isn't right somewhere. And it smells like hacking.

Especially, this phrase, "When started, the guy channeler out of the advisor (either crashed by mistake, or the UPU died), but the owner of TC is not even invited to monitor this situation".

It's easy to tell by the lifetime of the trades... in this case it looks like the time is breaking records.
 

I noticed that charts on the link in the first post are identical for almost all types of accounts, and there are a lot of them: ECN, STP, PAMM_ECN, PAMM_STP, Crypto (!).
Is there a mistake in methodology of calculation of these statistics?

 
Nikolay Sarjenuk:
This is easy to understand by the lifetime of the trades... in this case it seems that the time is breaking records.
The lifetime of the trades does not exceed a second. Maybe you mean order lifetime.

Also, in my opinion, it is more logical to judge not by the time the order is hanging, but by its effect on the spread and the moment of pouring => increase in spread.
 
Nikolay Sarjenuk:
It's easy to tell by the lifetime of the deals... in this case it looks like time is breaking records.
Maybe I'm wrong!
 
Running five or six (maybe more) copies of the same TS, but with different input options for diversification.

TS is a rollover channeler with MM as a percentage of balance (I haven't calculated the size yet). Each copy of TS always keeps one position open for ~18 lots. At the same time the take position hangs in the same place as the opposite limiter for the same volume. A flip, in short. That's why the gangs are doubled in volume - ~37 lots. These gangs are clearly visible in the screenshots for the 3rd, 9th and 12th.

When I run, you can see in the screenshot that the gangs have halved in volume.

Whereas previously the limiters were constantly modifying, these gangs were standing still as if they had been in place for more than a day. This suggests that the EA either crashed due to an error in the code or something happened to the terminal (UPU). I.e. the last thing that has been done by the TS itself - setting (modifying) Limits. And not empty, but with targets. Limiters worked, forming half gangs in the form of tees. And, of course, these tees hung without modifications of limiters (tees) usually inherent to this TS.

For more than a day hung - the owner did not monitor the account at all.

The whole point of TC is the algorithm of channel creation. It seems to be a good channel builder, so I made an offer.
Whoever has channelers and interest, get your TS on EURCHF 01.10.2015 - today and post your result. That way it will be clearer how much better the performance of the TS can be.
No one responded, so I decided to compare purely my handy with the guy in the field.

Free and publicly available is Level2-history, where you can get a good look at this guy, clearly assessing how much he's traded and, most importantly, how. Since the question of working with Level2-history is beyond the scope of metha, it is not touched upon here. But the potential is there - so it would be wrong not to mention it.

No clever analysis was done. It's all knee-deep and blunt. It's accessible by many people, so this code is just a demonstration of MQL4++ capabilities applied to rare - intelligent trading and re-selling. There is no hidden meaning, advertising or other conspiracy nonsense in this article. Writing helps you better understand your own thoughts. On the serious response is not expected initially.

If anyone is interested not on chit-chat/reading level, I suggest to post working code on Level2 research theme. You can simply download the history of the market and build, for example, BP of differences of total volumes on the ask and bid sides. Then overlay this on the SVR and check out the correlations.

The subject has nothing to do with the exchange. There are other laws here.

You are hooked on the channel guy. You can see that the man is actively engaged in business, not in chitchat and tidbits. I thought I would reach this guy by tricking him on different forums, such as forexcityems and forexfactories. To communicate in a constructive way. But practice shows that you just get bogged down in those bogs without any positive result. So the author is definitely not aware of public discussion of his trading.
 
zaskok3:
There are five or six (maybe more) copies of the same TS running, but with different variants of input parameters for diversification.

I got hooked on the channel guy himself. I could see that he was actively engaged in something useful, not just chattering and fluffing. I thought I would find this guy and called him on different forums, such as forexcasters and forexfactors. To communicate in a constructive way. But practice shows that you just get bogged down in those bogs without any positive result. So, the author is certainly not aware of public discussion of his trading.

I think this is a waste of time, if the guy is working with such lots, he must have a multi-billion account and be some kind of dealing centre representative.

So, he obviously has nothing to do with you!

 
EURCHF changed its characteristics (backtests of different channelers clearly show it) around September 22nd. At that time, judging by Level2-history, the guy was already trading. That is to say, my craft is a pale shadow of his channeler. Maybe he's still a grider a bit, but I doubt it.

Didn't do any additional analysis of the kid's TC, as it seems to be getting more use out of my own efforts so far. So far such a backtest:


Obviously a fit, despite the lack of OOS failure. There's no telling when it'll break.
 
zaskok3:
Obviously a fit, despite the lack of OOS failure. There's no telling when it will break.

And most importantly, at what cost.

Do you have feet? How big are they?

Reason: