In an effort to enhance the stability of Brawler System Trading, I recently attempted to incorporate low-correlation assets into my primary alpha logic.
To get straight to the point, I have decided to exclude this update from the live system for now.
1. The Strategy: A Safety Net for Market Crashes
The theoretical reasoning behind adding low-correlation assets was clear: I expected them to act as a safety net during sharp market downturns.
In fact, long-term backtesting showed a meaningful improvement in Maximum Drawdown (MDD).
Although overall returns decreased slightly, I believed the trade-off was worth it to increase the system's survival rate.
2. The Unexpected Obstacle: The Bid-Ask Spread
However, the critical mistake I made during the implementation phase was failing to precisely analyze the bid-ask spreads at the actual time of execution.
Existing Core Assets: These are highly liquid, with tight spreads maintained throughout all trading hours.
Newly Added Assets: While they offered attractive low correlations, the bid-ask spreads at the time of purchase were far too wide. Even during peak trading hours, the gaps remained significant.
3. Conclusion: The Law of Large Numbers vs. Strategic Retreat
The alpha logic of the Brawler System is designed to accumulate profits through frequent trading, relying on the "Law of Large Numbers." Given this structure, the slippage costs caused by wide spreads are likely to undermine the entire system’s profitability.
Ultimately, I concluded that it is better to exclude these assets altogether rather than absorb the losses incurred by poor execution prices.
4. Moving Forward
It is disappointing to see a promising idea hit a wall in reality. However, this experience served as a reminder of how crucial the "trading environment" is, beyond just backtest figures. For now, I will simply adjust the weights of those specific assets to zero. I look forward to testing new strategies and ideas again soon.
Comments
Post a Comment