Quantile Estimator

Overview

This indicator implements a robust statistical method for analyzing price distributions and detecting outliers in financial markets. It's particularly valuable because it:

1. Uses volume-weighted calculations for more accurate market representation
2. Handles non-normal price distributions effectively
3. Provides robust outlier detection through adaptive fencing

Key Components

Core Statistical Methods

• Harrell-Davis Quantile Estimator: Calculates weighted percentiles using beta distribution
• Quantile Absolute Deviation (QAD): Measures spread around the quantile estimate
• Double QAD Fencing: Creates adaptive upper/lower bounds for outlier detection

Main Parameters

• Sample Source: Default is HL2 (high-low average)
• Sample Size: Default 39 bars
• Probability: 0-100% (50% is median)
• Trim Percentage: Cuts low-density areas from calculations
• K-Multiplier: Controls fence width for outlier detection

Visual Output

• Blue line: Main quantile estimate
• Upper/lower bands: Outlier detection fences
• Shaded areas: Deviation regions

Comparison To Traditional Indicators

1. Distribution Agnostic
   • Works with any price distribution shape
   • No normal distribution assumption required
2. Robust to Outliers
   • Trimming reduces impact of extreme values
   • Adaptive fencing adjusts to market conditions
3. Volume Integration
   • Optional volume weighting for better market depth representation
   • Can operate without volume data if needed

Usage Guidelines

For Trend Analysis

• Monitor main quantile line direction
• Watch for sustained breaks of fences
• Use larger sample sizes for longer-term trends

For Volatility Assessment

• Track fence width changes
• Monitor frequency of fence breaks
• Compare upper vs lower deviation areas
  
  

Implementation Notes

1. Performance Considerations
   • Volume weighting adds computation time
   • Larger sample sizes increase calculation load
   • Double QAD mode requires more processing
2. Parameter Optimization
   • K-multiplier affects signal sensitivity
   • Trim percentage impacts calculation speed
   • Sample size balances responsiveness vs stability
3. Error Handling
   • Handles missing volume data gracefully
   • Maintains numerical stability with extreme values
   • Provides fallbacks for edge cases