The Hardest Decision
Building a system is hard. Running it is hard. But the hardest decision is recognizing when its time to stop.
Knowing when to kill your system is as important as knowing how to build it.
Legitimate Reasons to Stop
Edge Decay: Your edge has genuinely eroded. Markets changed, competitors arrived, or the pattern you exploited no longer exists.
Signs: Consistent underperformance vs expectations over extended period. Multiple validation methods confirm decay.
Risk/Reward Shift: The potential downside no longer justifies the upside. Maybe markets became more volatile. Maybe your life circumstances changed.
Signs: Sleepless nights. Stress affecting other areas. Willingness to accept losses changed.
Opportunity Cost: Better uses for your capital, time, or attention exist.
Signs: Other strategies with better risk-adjusted returns. Life opportunities requiring capital. Trading becoming a burden rather than opportunity.
Operational Burden: The work required exceeds the benefit.
Signs: Maintenance consuming too much time. Enjoying it less. Dreading system issues.
Illegitimate Reasons to Stop
Recent Losses: Drawdowns happen. Expected losses are not a reason to quit. Only quit if the losses exceed what your edge should produce.
Impatience: Profits take time. If youve validated properly, give the edge time to work.
Fear: Some fear is healthy. Panic after normal drawdown is not a reason to quit.
What Others Are Doing: Someone elses gains or losses are irrelevant to your edge.
The Kill Criteria
Define these BEFORE you need them:
Maximum Drawdown: If drawdown exceeds X%, stop and investigate.
Maximum Underperformance: If live results underperform backtest by more than Y% over Z trades, investigate.
Maximum Duration: If no new highs in N months, investigate.
Statistical Significance: If hypothesis test rejects the edge with 95% confidence, retire it.
Having predefined criteria removes emotion from the decision.
Investigation vs Termination
Hitting a kill criterion triggers investigation, not necessarily termination.
Investigation Questions:
- •Is this within expected variance?
- •Has something fundamental changed?
- •Is this execution problem or edge problem?
- •What does extended backtest show?
Continue if:
- •Investigation shows normal variance
- •No evidence of edge decay
- •Problems are fixable
Terminate if:
- •Clear evidence of edge decay
- •Fundamental market change
- •Problems unfixable at reasonable cost
Graceful Shutdown
If deciding to stop:
- •Stop generating new signals
- •Close existing positions orderly
- •Cancel pending orders
- •Document learnings
- •Archive code and data
- •Move capital to next opportunity
Dont rush the exit. Orderly shutdown prevents unnecessary losses.
Preserving the Learnings
Even failed systems teach lessons.
Document:
- •What worked and didnt
- •Why you think it failed
- •What youd do differently
- •Code and data for future reference
Reflect:
- •Was validation adequate?
- •Were expectations realistic?
- •What warning signs did you miss?
The Emotional Challenge
Stopping feels like failure. It isnt.
Reframe:
- •Protecting capital is success
- •Recognizing decay early is skill
- •Moving to better opportunities is smart
The traders who survive are the ones who exit losing strategies before they become disasters.
After Stopping
Options:
- •New edge development
- •Take a break
- •Trade existing systems only
- •Different markets or approaches
Stopping one thing creates space for something else.
Course Conclusion
Youve completed the School of Quant Trading. You now understand:
- •Why most strategies fail (Foundation)
- •What data you need (Data)
- •How to validate properly (Validation)
- •How to find edges (Discovery)
- •How to manage risk (Risk)
- •How to automate (Automation)
- •How to run in production (Production)
This knowledge positions you to either build your own system or evaluate others systems with informed skepticism.
The Path Forward:
Option 1: Build. Use this knowledge to construct your own trading operation. Its a significant undertaking but rewarding for those with the capability and time.
Option 2: Partner. Find services that implement these principles properly. Now you know what to look for and what questions to ask.
Either way, youre better equipped than 95% of retail traders who trade on hope rather than edge.
Welcome to quantitative trading. Trade well.