Duration tracking measures how long your equity stays outside the confidence bands. This single metric answers the critical question: "When should I stop trading this strategy?"
The Core Principle
By definition, your equity should be:
- Inside the bands 90% of the time
- Below the lower band 5% of the time
- Above the upper band 5% of the time
If you spend significantly more time outside the bands than expected, your strategy is not performing as the backtest predicted.
Key Duration Metrics
1. Total Time Outside Bands
Percentage of total bars spent outside the expected range:
| Time Outside | Interpretation | Action |
|---|---|---|
| 0-10% | Normal, as expected | Continue trading |
| 10-20% | Slightly high, monitor closely | Review recent changes |
| 20-30% | Concerning | Reduce position size by 50% |
| 30%+ | Strategy not working as backtested | Stop trading, investigate |
2. Longest Consecutive Streak Outside
Maximum number of bars spent outside bands without returning to normal range.
Example thresholds for daily timeframe:
- 10-20 days: Normal variance, happens occasionally
- 30-60 days: Extended period, review strategy parameters
- 60+ days: Major concern, likely regime change or strategy failure
3. Longest Streak Below Lower Band
This is the most critical metric. Extended time below the lower band indicates severe underperformance.
| Duration | Status | Action |
|---|---|---|
| 1-2 weeks | Within variance | Monitor but continue |
| 1 month | Concerning | Reduce position size |
| 2-3 months | Red flag | Reduce to 25% size or pause |
| 6+ months | Strategy failed | Stop trading completely |
Why Duration Matters More Than Single Points
Compare these two scenarios:
Scenario A: Brief Dip
Days 1-10: Inside bands Days 11-15: Below lower band (5 days) Days 16-30: Back inside bands Total outside: 17% (5/30 days) Longest streak: 5 days Action: Normal variance, no concern
Scenario B: Extended Underperformance
Days 1-5: Inside bands Days 6-70: Below lower band (65 days) Days 71-90: Slowly recovering Total outside: 72% (65/90 days) Longest streak: 65 days Action: Stop trading immediately
Both scenarios show time below the lower band, but Scenario B is catastrophic. Duration reveals the difference.
Setting Your Stop Rules
Conservative Approach (Recommended for New Traders)
30 consecutive days below lower band → Reduce size by 50% 60 consecutive days below lower band → Stop trading Return to bands → Resume at reduced size for 30 days Stable for 30+ days → Return to full size
Aggressive Approach (For Experienced Traders)
45 consecutive days below lower band → Review and adjust 90 consecutive days below lower band → Stop trading Multiple <30 day dips in 6 months → Investigate strategy health
Percentage-Based Rule
If time outside bands exceeds 20% of total trading days → Pause If time below lower band exceeds 15% → Stop trading Measure over rolling 90-day window
Real-World Example: The 10-Month Problem
You mentioned a strategy that stayed below the lower band from September to July (10 months). Here's the analysis:
Total visible period: 18 months Time below lower band: 10 months (55%) Expected time below: 5% (0.9 months) Actual vs. Expected: 11x worse Conclusion: This strategy failed completely. It should have been shut down after month 2-3 at the latest.
Why This Happened:
- Market regime changed: Conditions shifted in September, strategy couldn't adapt
- Overfitting: Backtest was too optimistic, didn't capture real variance
- Edge disappeared: Other traders found the same pattern, arbitraged it away
- Small sample backtest: Confidence bands were too narrow, didn't reflect true risk
Tracking Multiple Streaks
Don't just look at the longest single streak. Multiple shorter streaks are also a warning:
Pattern 1: Single Long Streak
Year 1: 1 streak of 90 days below lower band Total streaks: 1 Interpretation: One-time event, possibly recoverable
Pattern 2: Repeated Streaks
Year 1: 3 streaks (30 days, 25 days, 35 days) below lower band Total streaks: 3 Interpretation: Systematic problem, strategy is unreliable
Pattern 2 is worse even though no single streak is as long. The strategy repeatedly fails to stay within expectations.
Above Upper Band: Good or Bad?
Extended time above the upper band might seem good, but it's concerning:
- Overfitting indicator: Backtest was too optimistic, real trading will regress
- Lucky streak: Unsustainable run, expect reversion to median
- Market regime temporarily favorable: Will end when regime changes
- Risk of overconfidence: Might increase position size right before reversion
Spending 60+ days above the upper band suggests the backtest underestimated variance or you got very lucky. Don't expect it to continue.
Adjusting Position Size Based on Duration
Dynamic Sizing Formula:
Base position size: 100% If inside bands: 100% If outside bands 1-20 days: 100% If outside bands 21-40 days: 75% If outside bands 41-60 days: 50% If below lower band >60 days: 0% (stop trading)
Example Application:
Starting capital: $100,000 Base position per trade: $10,000 Day 30 below lower band: Trade with $7,500 (75%) Day 50 below lower band: Trade with $5,000 (50%) Day 70 below lower band: Stop trading entirely
Recovery Protocol
After returning to within bands, don't immediately go back to full size:
- Wait for confirmation: Stay inside bands for 10+ consecutive days
- Resume at 25% size: Test the waters with reduced risk
- After 20 days stable: Increase to 50%
- After 30 days stable: Return to 100%
- If you drop out again quickly: Extended break, strategy is unstable
Common Mistakes
- Only tracking max duration: Multiple shorter streaks add up, track cumulative time
- Not setting rules beforehand: Decide your thresholds before going live, emotions will cloud judgment
- Ignoring above-band time: Extended outperformance is also a red flag
- Not using a calendar: "60 days" feels different when you're living through it, track explicitly
- Hoping for recovery: If you hit 90 days below, stop hoping and stop trading
Automation Recommendations
Set up automated alerts:
- Daily email when outside bands (just for awareness)
- Warning after 30 consecutive days outside
- Critical alert after 45 consecutive days below lower band
- Auto-pause trading after 60 consecutive days below
When to Restart After Stopping
Don't rush back. Consider restarting only if:
- You've identified why the strategy failed
- You've backtested the strategy on recent data and it still works
- You've paper traded for 30+ days successfully
- You've implemented fixes or parameter adjustments
- Market conditions have returned to favorable regime
If you can't check all these boxes, the strategy is dead. Move on.
In VivaTrades
VivaTrades automatically tracks duration metrics on your equity curve when confidence bands are enabled. You'll see total time outside bands, longest consecutive streak, and a specific warning if you've been below the lower band for more than 60 bars. Use these metrics to set disciplined exit rules before going live.