It is I, Robopip, and I am back from my pod with the backtest results of the slightly revamped Simple System32.
As I pointed out in my previous post, I have previously backtested the system using its original settings that were specified by its maker, System32. However, upon getting very dismal results, I decided to backtest it again using a 50 simple moving average instead of the 100 SMA, in hopes that the faster simple moving average would help generate more signals. The results were slightly better… but only slightly:
As you can see, there were three trade signals that materialized in the system with the 50 SMA, just one more than what we got with the original settings! Two of those ended up as winners, but the other one incurred a nominally computed 150-pip loss which translates to a 1.00% hit.
Here is my verdict for the Simple System32 with the 50 SMA:
The system was still profitable over the past 8 months with a total return of 0.75%. But of course, I must take into account that despite the change to a faster moving average, the system still only had 3 trade signals. This would be perfectly fine if each trade led to an average of 10% gains per trade, but sadly, this is not the case.
Risk Tolerance: 6/20
Annualizing the results, I have come up with a Sharpe ratio of -0.4802, which is still below the Robopip Standard of at least 1.0000.
As I pointed out in my last post, the system does receive some points for its risk management rules, as it allows you to capture a decent portion of the trend via its scaling out technique.
I decided to leave this category unchanged, as the only difference is that a faster moving average was used.
Overall, the system is still confusing, as it includes applying indicators on an indicator, as well as using multiple moving averages, whose uses were not explained in the System32 thread.
In my home planet, we have a saying that roughly translates to: Best Things. Robo-life. Simple.
The system should be easy-to-understand and intuitive. It should be SIMPLE. The Simple System 32? Not quite.
Total Score: 17/50
A slight improvement, but far from being named the Holy Grail.
In my analysis, the main problem lies with the trend indicator, the 50/100 SMA. This is where all the wiring goes haywire and the system fails to generate enough signals. My assumptions are that the MA should be below price when all the indicators line up for a buy signal and an arrow pointing up materializes. But most of the time, the arrow is pointed downward while price is above the MA. The opposite happens for a downtrend.
The rest of the system appears to be fine, and could be the foundation of a decent discretionary system, but more than a few tweaks need to be made in order for it be used as a purely mechanical system.
In his defense, I believe that System32 does take a more discretionary approach in using this system and which is why he is able to catch more signals. In addition, Styngray and Coolncalm have gotten profitable trades forward-testing the system. If you’re interested in using the system, perhaps you can ask them how they trade it as it looks like System32 still hasn’t escaped from his abductors.
For now, the search for the Holy Grail continues. If you feel that you have a system that might just be the Holy Grail of mechanical systems, then please, submit it to the next Trading System of the Month contest!