The system is based on the correlation table below, and is in fact quite simple to understand and implement. The aim of this thread is to fine-tune the settings and develop a simple EA to automate the trading.
As you can see in the correlation table, certain pairs overreact to a counter-correlation price movement of two other correlated pairs, and always do it in the same direction. Iīve found out that, indeed, a strong movement of gbpjpy occurs when both gbpusd and usdjpy move in the same direction, with gbpjpy moving in the same direction as well. From these observations, it doesnīt really matter which currency, or currency pair, starts the movement. The fact is that gbpjpy moves quite strongly when gbpusd and usdjpy show an M15 bar moving in the same direction (when they should be moving against each other, since they are negatively correlated pairs). While such counter-correlation setting between gbpusd and usdjpy usually lasts for no more than one M15 bar, the price of gbpjpy moves in that same direction for at least another additional M15 bar, with a range quite larger than that of the two other pairs. Manual trades proved to be quite successful.
So, hereīs the system:
Timeframe M15
When gbpusd and usdjpy close a bar with the same color (both going north, or both going south), and both bars have a high-low range of at least 10 pips, enter a trade on gbpjpy in the same direction right at the opening of the new M15 bar.
SL and TP were tested with fixed values (30/20), yet they may be defined as a function of the first bar range.
Thatīs it.
Issues for discussion:
1. SL and TP settings
2. Minimum high-low range to enter a trade
Expert Advisor:
The simplicity of the system cries to be implemented into an EA. I am a programming illiterate, so help here would be much appreciated.
Required Inputs
SL
TP
Min range to enter (in pips, adapted to 4- or 5-digit brokers)
Lotsize
News filter, ideally
Trading hours
The same concept can be applied to other correlated pairs, as shown in the correlation table. I only checked the three pairs mentioned above. I have no idea which timeframe, minimum high-low range or SL/TP would be best for other sets of pairs, so it would need further analysis. An EA allowing the definition of the three pairs would be more versatile and will let us test any set of pairs.
Middleastern
As you can see in the correlation table, certain pairs overreact to a counter-correlation price movement of two other correlated pairs, and always do it in the same direction. Iīve found out that, indeed, a strong movement of gbpjpy occurs when both gbpusd and usdjpy move in the same direction, with gbpjpy moving in the same direction as well. From these observations, it doesnīt really matter which currency, or currency pair, starts the movement. The fact is that gbpjpy moves quite strongly when gbpusd and usdjpy show an M15 bar moving in the same direction (when they should be moving against each other, since they are negatively correlated pairs). While such counter-correlation setting between gbpusd and usdjpy usually lasts for no more than one M15 bar, the price of gbpjpy moves in that same direction for at least another additional M15 bar, with a range quite larger than that of the two other pairs. Manual trades proved to be quite successful.
So, hereīs the system:
Timeframe M15
When gbpusd and usdjpy close a bar with the same color (both going north, or both going south), and both bars have a high-low range of at least 10 pips, enter a trade on gbpjpy in the same direction right at the opening of the new M15 bar.
SL and TP were tested with fixed values (30/20), yet they may be defined as a function of the first bar range.
Thatīs it.
Issues for discussion:
1. SL and TP settings
2. Minimum high-low range to enter a trade
Expert Advisor:
The simplicity of the system cries to be implemented into an EA. I am a programming illiterate, so help here would be much appreciated.
Required Inputs
SL
TP
Min range to enter (in pips, adapted to 4- or 5-digit brokers)
Lotsize
News filter, ideally
Trading hours
The same concept can be applied to other correlated pairs, as shown in the correlation table. I only checked the three pairs mentioned above. I have no idea which timeframe, minimum high-low range or SL/TP would be best for other sets of pairs, so it would need further analysis. An EA allowing the definition of the three pairs would be more versatile and will let us test any set of pairs.
Middleastern