The satellite system: a second engine running beside the core
Adding a satellite system to my portfolio has successfully achieved the diversification I was looking for, though it functions more as a stabilizer…
Adding a satellite system to my portfolio has successfully achieved the diversification I was looking for, though it functions more as a stabilizer than a primary engine for growth. My goal with this second system was to isolate small, uncorrelated edges. I combined a daily Bollinger Band mean-reversion strategy with a seasonal filter (specifically restricting trades to April and December) and added a filter for USD trends limited to GBP/USD and USD/JPY.
Performance metrics
| Metric | Result |
|---|---|
| Annualized Return | 14.2% (over 11 years) |
| Monthly Average Return | ~0.10% |
| Max Drawdown | -8.3% |
| Profit Factor | 1.16 |
| Correlation with Core System | 0.29 |
| Prop Firm Step 1 | Pass |
| Failure Rate | 1.6% |
| The profit factor of 1.16 (which is the ratio of gross profit to gross loss) confirms the system is profitable, though the low monthly average of 0.10% means it is a slow grower. The correlation of 0.29 with my core system is the key takeaway here, as it proves I have successfully decoupled this strategy from my main breakout system. In other words, when one system hits a rough patch, the other is likely to act independently. |
Strategic role
While this satellite system is technically capable of passing prop firm challenges, its small scale makes it slow to reach targets on its own. It has a Monte Carlo pass rate of only 17.6%, which measures the probability of surviving prop firm rules based on simulated resampling of daily returns. Because of this, I position this system as a diversification tool for my core strategy. By running them together, the combined drawdown (the peak-to-trough decline in account equity) is reduced. This allows me to re-leverage the core system, effectively lifting the performance of the entire portfolio.
Lessons in filtering
I found that excluding general USD bidirectional trends is mandatory for this system to work. When I tested versions that included these trades, the drawdown ballooned to -25%, which would result in immediate disqualification from a prop firm. The success of this system relies entirely on isolating the positive, uncorrelated components and strictly stripping away the noisy, high-drawdown elements.