The 12-18% gap between Parametric and Historical VaR reflects the well-documented fat-tail problem in equity returns. For a $1B equity portfolio, this gap represents $12-18M in potentially unhedged tail exposure. Recommendation: use conditional (GARCH-implied) VaR for internal risk limits, Parametric for regulatory reporting, and Historical as the conservative benchmark. The Kupiec backtest confirms the Parametric model's breach rate exceeds the theoretical frequency.