read more:  https://www.mql5.com/en/blogs/post/751379 

Hi, this requires a completely different system of position accounting. This will not be the case in the master grid. But I will consider this wish in my future developments.

Hello. I bought the Master Grid EA and I am very happy with it. A great job and great respect to the programmer. The operation of the settings is very well explained. But beware: as already in the documentation, it is highly recommended to use and test the backtest function sufficiently.

I have a few questions about optimal operation... Is there any experience of how to let the grid area go step by step with the market development and also regularly delete the most distant orders manually and accept small losses, but thus keep the drawdown low...? Another thought is to stop and restart the EA in small periods of time (daily, weekly, etc.).

How do you optimize setting? Whats the methodology?

I think there are many approaches... The first thing I did was define a possible range that the market could make. Then I looked at the stock value, in which frame of movement this stock value moves in market phases. To do this, I then divided the grid into the possible total range. Next, I assumed the worst-case scenario of open negative positions and calculated the margin to determine the possible lot size to match the account size.

I inserted these values into the Strategy Tester and then had the calculated values tested in a min and max range. It is important that you determine values that do not burden your account or margin too much. It's a bit of a gimmick, but the strategy tester is perfect for it. Learn to use it properly and it will be your GameChanger.

But as I said..., there are many ways... 

Hello. Thanks for your feedback! Next I will answer point by point, but as you yourself noted, this is just my opinion based on my own experience:

1. Closing unprofitable positions during long-term trading (in my opinion) depends primarily on what currency instrument is being traded. For example, if trading is on XAUUSD and there are old unprofitable ones below the level of a year ago, I think they can be safely closed. Because no matter how much gold fluctuates, in fact it always grows. But if trading takes place, for example, on EURUSD, then closing unprofitable positions is not at all obvious, even if they are far from the current market level.

Another thing is important here... the question is how much money is lost on swaps (!) of unprofitable positions. If you can find an account without swaps (and brokers may call it something else), then the only thing that matters for long-term trading is whether you have enough funds to withstand a drawdown. And this can already be dealt with in a variety of ways, from locking to increasing the total market lot in the direction of the trend.

If the account has swaps and they are large enough, then yes, it is better to close unprofitable positions at some point.

2. Short-term trading undoubtedly also has a right to exist. But here it is important to take into account another point, which concerns grid trading. In most cases, grid trading starts with losses. That is, if the idea is to keep price levels on either side of current market prices, there will be a constant loss hanging in the profit column. But if the flat continues, even for some time, and the take profits and other settings are chosen correctly, the final profit may exceed this loss many times over. This moment is very convenient to track using the trading mode panel with profit tracking.

My method is to actively use the tester's optimizer function. That is, I find a period in history similar to the one that I predict for the current moment and launch the advisor on it in the mode of optimizing the parameters I have chosen. The more parameters and the historical testing period, the longer the optimization takes. As a result, I look at which settings best suit my goals (in terms of profit, drawdown, various odds, etc..). Then the selected settings can be tested on another similar historical period, and if they are again successful, you can launch the advisor with them in real trading.

This approach can be used for both short-term and long-term trading. And this does not mean at all that the settings should remain unchanged. One of the main ideas of my grid advisors is precisely that almost any settings can be changed at any time. That is, if you see that the market is changing (or your opinion about it), try to select the settings for another similar historical period and put them into real trading.

p.s. To make optimization faster, I recommend choosing the simulation mode: only opening prices
The accuracy in this case can be adjusted by the time-frame.

Thanks

Thanks

Thank you for your explanations. I'm going to implement this like this...

Another question about the grid: is it possible to go e.g. only in the long direction, but to open the second long position only when the previous one is closed. So to follow a trend and take profits in steps, the risk remains limited if the trend ends and the last long lot is not filled?

Perhaps I didn't understand the question entirely correctly, but what’s the point of closing one position and immediately opening a second one in exactly the same direction?

This can be done using several charts of the same currency instrument and on each of them you need to launch an advisor with a different magic number and its own settings.

But doesn't a grid of pending orders with take profits of the right size do exactly the same thing? The market position is closed, but at the same price there is a pending order, which immediately becomes a market one.

Roughly speaking, yes. However, with the disadvantage that several positions have to be opened and closed within a trend and thus possibly cause more costs. On the other hand, an order that is opened at the beginning of a trend and can 'take' the entire trend with a trailing stop would only have to open a position once.

The idea behind this is when you recognize a trend, e.g. in higher timeframes, open a position and start it according to defined trailing stop rules.

Why several positions, set up the grid so that there is always only one market position on the market (according to the principle that I described above). Moreover, you can even check how it all works in the tester without having to spend additional time, money and nerves on it.

If it were possible to clearly identify the beginning and end of a trend, there would no longer be any need for other strategies.

At the moment, the idea you described remains incomprehensible to me personally. And so far I don’t understand how a trailing stop can help here. At the same time, my personal experience with using a trailing stop is that it almost always only reduces potential profit.

How can I set up the grid so that only one position is opened within a defined area? Would this work if you set the grid to 10000 and the take profit to 1000 as an example?

In this case, would a new position be opened again and again as soon as the first position was closed at a lower profit target?

For example for EURUSD:

BuyUp orders only

Initial price: 1.07

Step: 1000

Take Profit: 1000

Stop Loss (at will): 1000

This example does not take into account the spread, this is only to show the principle of how only one market position can work in the grid.

In this case, the restoration of closed positions can be disabled in the settings (for all Buy orders)