Extreme Back Testing Regime
At GUARDIAN FX ENGINE, we invest hundreds of hours into refining each individual trading strategy to ensure peak performance. And then we do it again, and again...
​
Our system comprises multiple, specialised trading bots, each designed for different market conditions. While a single, all-encompassing bot isn't feasible, combining targeted strategies enhances overall performance and increases diversity, potentially leading to substantial profits.
Each strategy has to pass a major, comprehensive testing regime to be admitted to GUARDIAN FX ENGINE.
Strong individual results for each strategy are essential, but ensuring they work cohesively as a portfolio is equally important. We aim to avoid strategies that trade similarly, as this would reduce our diversity. Diversity is crucial for maintaining steady and stable growth.
Before any strategy is included, it undergoes rigorous portfolio testing. This comprehensive analysis evaluates how strategies perform together in our back testing environment. We often discard even solid strategies at this stage to preserve portfolio diversity and enhance overall performance.
​
Below you can see how comprehensive our development and testing is.
Testing Regime
Overview of Example Individual Strategy
Currency Pair: GBPUSD
Time Frame: M5
Back testing return per year: 9%
Max draw down (on back test): 1.13%
Winning Percentage: 94%
Expected number of trades per year: 145
Money Management: 0.5% of balance
Testing Range: 01/01/22 to 19/05/2024
Overview of Whole Portfolio of Strategies
Equity Growth Chart of Individual Strategy
Equity Growth Chart of Portfolio
Comprehensive Trade Analysis of Individual Strategy
Comprehensive Trade Analysis of Portfolio
Monte Carlo Testing
What is Monte Carlo Testing?
​
Historical results of a trading strategy give us some prediction of the future performance. By using Monte Carlo analysis you’ll be able to make this prediction much more accurate.
​
The basis of Monte Carlo method is running the same simulation a number of times, each time with small random changes. The higher the number of repetitions, the bigger is the statistical significance of the results. In these tests, we do small, random changes in many parameters including the order of trades, missing trades and various other small adjustments to the trading criteria.
​
We do this 100 times as this gives as much greater confidence for the future.
Monte Carlo Testing of Individual Strategy
Monte Carlo Testing of Portfolio
Cross Checking
We also back test on a different platform to ensure that we are getting accurate and corroboritive results.
Back Test On Metatrader 4 (99% quality tick data)
Portfolio Correlation
We don't want our strategies to be similar. If we did, it would be just like having one strategy and trading at larger trade sizes. This would decrease diversity and increase volatility (eg. larger drawdowns)
To ensure we have high diversity, we run them through a test to see how correlated they are. We want low correlation and this is highlighted as green
