Structure rules. Learning how to structure programmes, exploring possibilities, errors, solutions, etc. - page 15

New comment
 
GaryKa:

The point is that the starter most likely has not a strategy, but a price predictor.

A correction here. The predictor is not the price, but the direction of price movement. Otherwise I agree.

And what you(C-4) are talking about is the work of a Money Management module, which receives both predictor readings and past trade results as an input (some kind of a function). If there is no MM, the final trading algorithm essentially converts the virtual predictor position (which could not care less about past results of trading) into a real one, where the future market trend is a sign of a recommended position, and confidence/probability is proportional to the volume of the same position.

MM module is a layer between the predictor and driver and it works in different ways, from capitalization and risk limit (relative drawdown in X ... hours/trades/pts of movement) to a radical reversal of the position recommended by the predictor.

You can interpret it this way too.
 
MetaDriver:

These robots definitely need treatment. The treatment can be very simple. All it takes is motivation. This is basic.

And to be motivated, they need to see and appreciate the colossal superiority of netting thinking over order thinking. As long as they're sick, I won't let them anywhere near the healthy population - let them sit in quarantine...

:)

What isn't? Do I have to apologize? Or am I guessing correctly? ;-))

No, you haven't:) In my model there is no concept of remembered signal. It's simpler, because A robot can do two and only two things. Either open a position or close it. There is no third option. A position may be either long or short. If there is a long position, it is checked against the long close conditions; if there is a short position, it is checked against the short close conditions. The scheme is very simple:

//Событие, наступил новый бар
protected void OnNextBar()
{
   //Проверяем условия открытия длинной позиции
   InitBuy();
   //Проверяем условия открытия короткой позиции (последовательно неважна)
   InitSell();
   //Если есть короткая позиция, проверяем ее на условия закрытия
   if(ShortPosition.Active)
      SupportSell();
   //Если есть длинная позиция, проверяем ее на условия закрытия
   if(LongPosition.Active)
      SupportBuy();
}

That's it! No one remembers anything, it only works with what it has at the moment.

 
GaryKa:

The point is that the starter most likely has not a strategy, but a price predictor.

And what you(C-4) are talking about is the work of a Money Management module, that receives both predictor readings and past trade results as an input(some kind of function). If there is no MM, the final trading algorithm, in fact, will transform a virtual position of the predictor (who doesn't care about past results of trading) into a real one, where the future market direction is a sign of a recommended position, and the confidence/probability is proportional to the volume of the same position.

MM module is a layer between the predictor and driver and it works in different ways, from capitalization and risk limit (relative drawdown in X ... hours/trades/pts of movement) to a radical reversal of the position recommended by the predictor.

MM is a change in volume. A flip (any flip) has nothing to do with MM.
 
MetaDriver:

These robots should definitely be treated. The treatment can be very simple. All it takes is motivation. This is basic.

That's a lot of robots to treat. Or maybe it's just that the scheme is not universal and not suitable for all cases.
 
C-4:

That's a lot of robots to treat.

I don't have to, I make the healthy ones straight away. And the thing is, they're easier to make than the sick ones. You can check. Just need a paradigm shift. That's where it gets stuck.


Or maybe it's just that the scheme is not universal, and not suitable for all cases.

I love democracy, but the truth is dearer. The scheme is even more universal than it seems at first sight.)
 
I have an excellent understanding of netting (if you don't praise yourself, no one will:))), I even published an article on the problems of multi-experts at one time. For several years now I've been trading on RTS, which is all about netting. In addition to MetaTrader I work with several other purely stock terminals (Quick does not count). So? So, these exchange terminals have a simple concept of position. There can be many of these positions, and they can be directed in different directions. There is no right and wrong thinking. There is comfortable thinking, and that is what must win!
 
C-4:

No, guessed wrong:) There is no concept of a memorised signal in my model. It's simpler, because a robot can do two and only two things Either open a position or close a position. There is no third option. A position may be either long or short. If there is a long position, it is checked against the long close conditions; if there is a short position, it is checked against the short close conditions. The scheme is very simple:

That's it! No one remembers anything and only works with what is currently available.

I won't play along. ;)

Where do we put all this stuff:

How does it not matter!? Yes in any strategy at the level of its logic, its current state is always known! Let's take a simple strategy on the intersection of two averages: it has only two states, it is either buying or selling. Without memorizing its position, it will open a long position every time it sees that the fast average is above the slow one. So what's the synchronizer to do? Say to her: "no, you already have a long position, I won't let her open another one!

In this case we will have to store the history of triggered signals somewhere, which is very expensive. Let's again consider crossing of 2 averages. Suppose we restarted the Expert Advisor. There is no new cross for entry and the EA will somehow need to restore its trading history and understand that there was a crossover and that it should be in the Buy state and that the signal has been processed and we should not open a new position, but we will need to find the old position, but it will not be easy to find, because the current position may not necessarily belong to only one EA... All in all a nightmare nightmare...............
That's all very well of course, but what about strategies whose current "recommendation" depends on a previously opened position. Suppose the strategy is actively pyramiding and has this condition (pseudo code):

:-)

I responded to all these arguments. And now it turns out that all this didn't happen at all?

And I have all the moves written down...!

;-)))

 
C-4: MM is a change in volume. MM does not refer to flip (any flip).

MM is money management.

  • - if you had 5 contracts and now have 2. then you have reduced the position by 3 contracts - that's MM. yes
  • - If you had 3 contracts and now you have -1. Then you reduced the position by 3 contracts - that's MM. ????? Yes. That's MM. Was there a flip? Yes there was.

... and its results can be anything, ... up to a radical reversal of the position recommended by the prognosticator.

See, the parser/prognosticator/brain/neighbour recommends buying assets. Your MM module remembers (has accumulated statistics) that these recommendations have been failing a lot lately (already breached the trust threshold). It makes sense to do the opposite, it does. It's just that MM's function is a bit more sophisticated, like investor behaviour.

P.S. It is difficult for a person to switch to netting; he is afraid and does not understand how it happens - opening and closing of orders, take profit and stop loss, the number of deals - everything disappears. Only one position remains (one number hovering near zero), and the need to maintain it. Plain and simple.

 
GaryKa:

P.S. It is difficult to transfer a person to netting, he is scared and does not understand how, opening and closing orders, take profit and stop loss, the number of deals - everything disappears. Only one position (one number hovering near zero) remains and we need to maintain it. Plain and simple.

++
 
GaryKa:

MM is money management.

  • - If you had 5 contracts and now you have 2, you have reduced your position by 3 contracts - that's MM .It's not clear whether thereduction was due to the money management formula or not. This is why it is impossible to answer definitively.
  • - If you had 3 contracts and now have -1. You decreased your position by 3 contracts. 4 contracts is MM. ????? Yes, that's MM. Was there a flip? Yes, there was.

See analyser/prognosticator/brain/neighbour recommends buying assets. Your MM module remembers (has accumulated statistics) that these recommendations have been failing a lot lately (already breached the trust threshold). It makes sense to do the opposite, it does. It's just that MM's function is a bit more sophisticated as an investor's behaviour.

P.S. It is difficult for a person to switch to netting, he is scared and does not understand how, opening and closing orders, take profit and stop loss, the number of deals - everything disappears. Only one position remains (one number hovering near zero), and the need to maintain it. Plain and simple.

No, no and again no!!!!!!!!!!!!!!!! By that logic, a simple strategy flip on a moving average is also MM. Why, it was +1 contract and became -1.
1...891011121314151617181920
New comment