As an ongoing feature of this website we will describe the approaches taken and showcase the performance results of a number of Programs that we are currently testing. For those who are interested and want to follow along with us, you can monitor live performance by using the FX Blue link for the relevant program.
We use a live account with a minimal $10K commencing balance to test each of our Programs to ensure the algorithms are free of execution issues prior to implementing with significant capital at risk. A standard opening balance applied to all our Programs also allows us to easily compare and contrast our various Programs.
The ATS 1000 program comprises 143 separate divergent systems (Expert Advisors) operating across 12 Forex markets on the daily timeframe. This diversified systematic program operates on the MT4 platform and has been scaled for a $10,000 opening balance.
Each system has a divergent design logic preset into it’s configuration which comprises at it’s heart a standard entry 50/200 SMA moving average crossover and a standard momentum indicator filter applied to the daily timeframe. In addition each system possesses an initial and trailing stop condition with no profit targets or other exit condition. This broad design logic ensures that we cut profits short and let profits run and allows for the generation of medium to long term divergent momentum/trend following EA’s that possess positive skew.
Each system has an additional entry criteria which is data mined to allow for uncorrelated design relationships, allow for a degree of system adaptability in response to emerging market conditions and ensure each system possesses unique risk weighted return signatures.
A standardized workflow process is used for data mining purposes and the portfolio is refreshed at annual intervals.
This system will remain active and continue to be monitored provided that designated Max Drawdown thresholds are not met. We allow for a max drawdown of 40% for all our Programs before we elect to turn them off. Future expected returns are unknown given the unpredictable nature of divergent market conditions, but based on backtest results, we are conservatively targeting a medium to long term future risk-weighted MAR ratio of 1.0.
Salient performance metrics of this program displayed in the table above are as follows:
- The Program backtest range spans the period from 1 Jan 2000 to current day and comprises 5,183 trades and trades in both the long and short direction;
- Lot sizes for each portfolio expert in the program are scaled for a $10,000 opening balance account for the Program.;
- The average holding period per trade is 39.6 days with a maximum holding period of 1,640 days (4.5 years). The program is a medium to long term trend following/momentum program;
- The Pwin% is 51.8% with an average risk to return ratio of 0.43 : 1 and a profit factor of 2.51;
- The compound average annual growth rate of the Program is 13.9% with a Maximum drawdown over the period of 8.8%; and
- The Program has high positive skew and a very healthy MAR ratio.
It is very important to note that future risk-adjusted returns are expected to be significantly diluted when compared to backtest results. This is in part attributed to the data mining methods deployed in the workflow process which naturally bias the results towards optimal performance results but also due to material differences between assumptions used in backtesting and those applied to the live environment and the obvious inability to predict future risk -weighted returns given an uncertain future. It is therefore prudent to significantly dilute our future performance expectations in light of these considerations. The backtest is therefore more an upper limit performance expectation as opposed to a realistic measure of future performance.
We avoid using the backtest as a basis to predict future expected returns but rather use it as a road-map to guide us into an uncertain future by providing us with tolerance thresholds. We use the backtest as a method to risk-weight the portfolio based on historic returns and to develop critical performance metric thresholds which we use to compare and contrast against actual live performance.
The return streams of the individual market contributors to the Program are as follows.
Overall Program performance for the backtest range is reflected in the equity curve below based on a $10,000 account.
The Program commenced on 17 June 2019 and no monthly performance results are currently available. To view live trading results, refer to the FX Blue link for this program at the commencement of this post. You will need to set the date filter to commence from 17 June 2019.
Well that’s a wrap guys.
Trade Well and Prosper