The relevance of expected shortfall models for hedging strate-gies in different time windows size

The paper investigates the relevance of Expected Shortfall (ES) models for hedging strategies across different time windows size. Exchange rate data for the Brazilian Real (BRL) against USD, EUR, JPY, and CNY were analyzed over two periods: 2000–2024 and 2020–2024. The analyses carried out to evaluate these indicators covered the period from August 1, 2024 to August 1, 2025, where it was found the exchange rate volatility of the Brazilian Real that reached the ES risk zone. All three types of methods—parametric, non-parametric, and semi-parametric—show distinct behaviors depending on the period evaluated. The findings reveal that non-parametric and semi-parametric ES estimators are more responsive to recent market changes than parametric ES estimators. The selection of appropriate time windows and estimation techniques are highlighted by the frequency that the ES breaches the limits of the volatility.