About this question
Easy · risk_management · Quant Researcher interview question · var_backtesting, lopez_score, model_selection, scoring_rule, risk_management
Value-at-Risk (VaR) model evaluation requires assessing not just the frequency of exceedances, but also their magnitude. The Lopez (1998) quadratic scoring rule provides a loss function that penalizes larger VaR violations more severely than smaller ones. This score is crucial in quantitative risk management for comparing and selecting the most appropriate VaR models. Task Implement the function lopez_var_score(returns: list, var_forecasts: list) -> float to calculate the average Lopez (1998) q