Interview, Podcast
The Shift to Systematic Trading — Building Backtested Confidence · Dave Mabe
- Human Emotion in Automation: Automated trading is not immune to emotion; the most difficult psychological hurdle is maintaining confidence during drawdowns, which is when traders typically quit rather than at all-time highs.
- Career Trajectory: Dave Mabe, a North Carolina-based trader with over two decades of experience, transitioned from the dot-com era swing trading to day trading around 2005, eventually moving from a discretionary approach to a fully systematic, rule-based gapping breakout framework.
- The Discretionary Limitation: Mabe discovered that while his manual execution had intuition, it failed to consistently filter trades; he realized that experienced discretionary traders often differ in results not because of strategy, but because of their subconscious ability to skip specific setups, a skill difficult to replicate mentally.
- Backtesting Shock: The first automated backtest Mabe ran of his manual strategy outperformed his actual live trading results, revealing that his "intuition" regarding trade selection was inferior to a systematic, rule-based filtering process.
- Scaling Metrics: Upon implementing automation, Mabe scaled his business "wide" (increasing trade frequency) rather than "up" (increasing position size per trade), jumping from approximately 300 trades in one year to 3,000 trades in the next, resulting in roughly five times the profit.
- Drawdown Management Failure: During a significant drawdown, Mabe initially reduced his position size to preserve capital, which caused the drawdown to deepen and nearly forced him to quit; he subsequently decided to maintain size through the recovery to avoid the lag time required to rebuild it.
- Confidence Building Process: Mabe advocates for a gradual transition to automation (e.g., automating position sizing first, then entry, then exit) and utilizing "forward testing" with tiny size to reconcile backtest discrepancies and build confidence before scaling capital.
- Systematic Filtering Rules: The core gapping strategy involves identifying stocks gapping up, waiting for a narrowing range post-open, and entering on a breakout with a stop placed at the opposite side of the tight range.
- Backtesting Philosophy: Backtests should serve as a "pristine" target performance rather than an attempt to model every real-world friction (like slippage or borrow costs) immediately; traders should account for these losses post-trade to avoid the "fool's errand" of over-optimizing simulation precision.
- The "Sidewalk" Analogy: Mabe uses a university construction analogy to argue that traders should deploy systems live immediately at small sizes to observe natural "wear paths" (failure points) rather than spending excessive time perfecting a theoretical backtest.
- Common Backtesting Errors: Mabe debunks the common rule of "taking half profits and moving stops to breakeven," noting that backtests show this significantly underperforms holding full positions to natural exits; similarly, he warns that tight stops in backtests often look unrealistic because bar precision can mask immediate stop-outs.
- Diversification Necessity: Mabe emphasizes that the hardest challenge for a successful trader is generating a second, uncorrelated strategy; relying on a single system creates a "fossilized" mindset and leaves the trader vulnerable to strategy decay.
- AI and Code Generation: Mabe utilizes Large Language Models (LLMs) like ChatGPT to generate and audit code, arguing that a basic understanding of programming logic is required to prevent "spaghetti code" and ensure the AI executes the intended strategy correctly.
- Forward Testing Utility: Forward testing is defined as executing live trades with minimal size to create a feedback loop for reconciliation, serving as a primary method to gain confidence alongside historical backtesting.
- Software Recommendations: Mabe recommends Amabroker, Trade Ideas (noted for not requiring code), and RealTest (limited to daily data) as tools for backtesting, offering a free course via betterbacktesting.com for beginners.
- Shorting Challenges: While acknowledging that shorting strategies often suffer from overlooked costs like locate availability and borrow fees, Mabe suggests keeping backtests "pristine" and accounting for these costs as a post-trade reality check rather than attempting to model unpredictable variables.