Historical VaR Percentile

Historical VaR Percentile represents the quantile-based risk measure computed from actual historical price movements and returns. For a quant scoring platform focused on insider trading detection and risk assessment, this metric reflects tail risk by identifying the actual loss threshold below which a specified percentage of observed historical returns fall. Unlike parametric VaR (which assumes normality), this non-parametric approach captures fat tails and regime-specific volatility clusters inherent in real market data, making it particularly valuable when analyzing concentrated or illiquid positions held by insiders or related parties.

In insider-trading surveillance, Historical VaR Percentile serves dual purposes: first, it establishes a baseline risk profile for a trader's or PDMR's typical position sizing and portfolio behavior under stress, and second, it flags anomalies when sudden spikes in concentration or drawdown depth exceed historical norms. The percentile-based approach avoids the model risk embedded in parametric assumptions, which is critical when dealing with irregular trading patterns, blackout window violations, or Section 16 reportable transactions where market conditions may deviate sharply from historical distributions.