Grid Hero Behind The Scene - "In-Sample" & "Out-of-Sample" Development And Testing Methodology
Let me reveal a little more about how Grid Hero was developed :)
Grid Hero's development phase was segregated into 2 phases call the "In-Sample" phase, and the "Out-of-Sample" phase.
The "In-Sample" phase is the primary focus of the EA's development where the EA's trading logic, algorithm, stability, profitability, stagnation, etc, are being developed and tested using a pre-defined test period (usually a few years).
Whereas the "Out-of-Sample" phase is a test period that falls outside of the "In-Sample" period, where the EA is being tested for Robustness, to see if it can exhibit the same performance in a test period that falls outside of the EA's development period.
When I developed Grid Hero, I used a "Reversed Sampling" methodology to define the "In-Sample" and "Out-of-Sample" period:
For In-Sample period: I used the late years of 2012 to 2017 because I prefer to use recent market data in establishing Grid Hero's foundation
For Out-of-Sample period: I used the earlier years of 2004 to 2011 to test Grid Hero's robustness in remaining profitable in a segregated period that falls outside of Grid's development phase.
The above methodology of using recent data as "In-Sample" period while using the early data as "Out-of-Sample" period is call "Reversed Sampling".
As evident in the back tests, Grid Hero is able to maintain its consistency in remaining profitable and stable even in the "Out-of-Sample" test period, which is a good form of robustness check.
And needless to say, all my tests are conducted with Real Tick Data (99% Modelling Quality), Real Variable Spread with Commission, and even with Execution Delay Slippage Simulation, as well as Monte Carlo Simulation Runs of Trade Randomization.
So I hope this little behind the scene story of Grid Hero gives you more insight of how much work and effort have been put into the development of Grid Hero EA. It is not just another EA that was developed out of randomness, but rather an EA that was carefully crafted and created using proper development and testing methodology.
Last but not least, stay tuned for the upcoming Self Adaptive Grid Hero Version 7...
GRID HERO V5 13 Years Back Test Video: https://youtu.be/QuWqw2OY8ps
GRID HERO V6 Visual Back Test Video: https://youtu.be/aBcesp-OTsU
GRID HERO EA Product Page: https://www.mql5.com/en/market/product/20937
GRID HERO EA FAQ: https://www.mql5.com/en/blogs/post/689594
GRID HERO EA Live Signals: https://www.mql5.com/en/signals/author/daveng
GRID HERO Myfxbook(new signal): myfxbook.com/members/iamforexrobot/grid-hero-real-axitrader-ecn/2195120
GRID HERO Myfxbook(retired): myfxbook.com/members/iamforexrobot/grid-hero-real-axitrader-ecn/2005396
Let me reveal a little more about how Grid Hero was developed :)
Grid Hero's development phase was segregated into 2 phases call the "In-Sample" phase, and the "Out-of-Sample" phase.
The "In-Sample" phase is the primary focus of the EA's development where the EA's trading logic, algorithm, stability, profitability, stagnation, etc, are being developed and tested using a pre-defined test period (usually a few years).
Whereas the "Out-of-Sample" phase is a test period that falls outside of the "In-Sample" period, where the EA is being tested for Robustness, to see if it can exhibit the same performance in a test period that falls outside of the EA's development period.
When I developed Grid Hero, I used a "Reversed Sampling" methodology to define the "In-Sample" and "Out-of-Sample" period:
For In-Sample period: I used the late years of 2012 to 2017 because I prefer to use recent market data in establishing Grid Hero's foundation
For Out-of-Sample period: I used the earlier years of 2004 to 2011 to test Grid Hero's robustness in remaining profitable in a segregated period that falls outside of Grid's development phase.
The above methodology of using recent data as "In-Sample" period while using the early data as "Out-of-Sample" period is call "Reversed Sampling".
As evident in the back tests, Grid Hero is able to maintain its consistency in remaining profitable and stable even in the "Out-of-Sample" test period, which is a good form of robustness check.
And needless to say, all my tests are conducted with Real Tick Data (99% Modelling Quality), Real Variable Spread with Commission, and even with Execution Delay Slippage Simulation, as well as Monte Carlo Simulation Runs of Trade Randomization.
So I hope this little behind the scene story of Grid Hero gives you more insight of how much work and effort have been put into the development of Grid Hero EA. It is not just another EA that was developed out of randomness, but rather an EA that was carefully crafted and created using proper development and testing methodology.
Last but not least, stay tuned for the upcoming Self Adaptive Grid Hero Version 7...
GRID HERO V5 13 Years Back Test Video: https://youtu.be/QuWqw2OY8ps
GRID HERO V6 Visual Back Test Video: https://youtu.be/aBcesp-OTsU
GRID HERO EA Product Page: https://www.mql5.com/en/market/product/20937
GRID HERO EA FAQ: https://www.mql5.com/en/blogs/post/689594
GRID HERO EA Live Signals: https://www.mql5.com/en/signals/author/daveng
GRID HERO Myfxbook(new signal): myfxbook.com/members/iamforexrobot/grid-hero-real-axitrader-ecn/2195120
GRID HERO Myfxbook(retired): myfxbook.com/members/iamforexrobot/grid-hero-real-axitrader-ecn/2005396
