Let's implement the grail net.

New comment
 
Above and below from the price with a step - delta1 you set the limit orders, at a distance from each other with a step - delta2. Put the lot size into the variables. Take equals delta2, we do not set the stop loss. When closing a position by Take order, an identical order is placed instead of the one that failed. Everything is cleared when the global take (balance + equity) is reached and the grid is set again. The lot increment coefficient is put into variables, i.e. the order values increase by this coefficient which is proportional to the distance from the price. Unfortunately, I am not a programmer. Then I want to continue: to combine this grid with the trend one and enter both of them not abnormal, not indicative, but sensible, based on instrument volatility + Fiba (to make the step correct). Someone do it, let's continue.
In the settings we should limit the trading depending on the days of the week.
 

What is a "trend grid"?


And what is "entering with sense and purpose"?

 
sever29 >>:
Сверху и снизу от цены с шагом- дельта1 выставляете лимитные ордера, на растоянии друг от друга с шагом- дельта2. Выносим в переменные размер лота. Тейк равен дельте2, стоплосса нет. При закрытии позы по тейку, на место отработавшего ордера выставляется идентичный ему. Все сноситься при достижении глобального тейка (баланс+средства) и сетка устанавливается снова. Выносим в переменные - кооф. увелечения лота, т.е. обемы ордеров увеличиваются на этот кооф. пропорционально удалению от цены. К сожалению я не проггер. Затем продолженние: совместить эту сетку с трендовой, а вход у обеих сделать не отбалды, не идикаторный, но со смыслом и толком, завязанный на волатильности инструмента+ фиба (для корректности шага). Сделайте кто-нибудь, продолжим...

Are there any results from working with such a system? That's what's stopping me=))

Take is equal to delta2, no stoploss.

 
I call a trend net the one that earns in a strong unidirectional movement (trend), while a flat net, on the contrary, both in a flat and a sluggish trend.
 
Necron писал(а) >>

Are there any results from working with such a system? That's what's stopping me=))

I can only give you a two year test on my robot, but it doesn't work as I'd like it to, but the gist is the same

Files:
testergraph.rar  27 kb
 
sever29 >>:
я называю трендовой сеткой, ту которая зарабатывает в сильном однонаправленном движении (тренде), а флетовая, наоборот, как во флете, так и вялотекущем тренде.

What does it look like?

 
In a "trend" situation, a grid of stop orders is placed; in a "flat" situation, limit orders are placed.
 

And how do we know which is trending and which is flat?

 
We don't need to know that, working with limit orders, due to the cof. lot, the robot can withstand trend movements, but strong 1,000p reversal movements cannot. By the time these movements have the beginning and the end (look at the history), it means that in case of a critical strong unidirectional movement we enter the lot using the "trend" grid. We sit out the "storm" and then go out of the lock in a flat, or in the same trend, but not so strong. In general I have a whole system and on the history there are certainly some unpleasant days when the drawdown will be significant (500 points in a day and a half)
 
sever29 >>:

могу только тест за два года выложить, по своему роботу, но он работает не так как хотелось бы, но суть таже

The curve is impressive, why the drawdown at the end if it's no secret=)

 
Necron писал(а) >>

The curve is impressive, why such a drawdown at the end if not a secret=)

the test was before a relatively strong unidirectional move, a day later, in a pullback, all would have closed in the +

1234567
New comment
Reason: