Quantile-based modeling of scale dynamics in financial returns for Value-at-Risk and Expected Shortfall forecasting

Battle Count: 65

Highly relevant for quantitative trading as it provides robust, distribution-free VaR and ES forecasts critical for position sizing, risk limits, and portfolio management. The method's ability to capture leverage effects, skewness, and heavy tails makes it particularly useful during periods of market stress. Directly applicable to regulatory capital requirements (Basel III) and internal risk monitoring. The QbSD-gAS specification consistently outperforms benchmarks in empirical applications across major international indices.

Moderate complexity. The method involves restricted quantile regression with non-crossing constraints, requiring constrained optimization. The sequential estimation strategy (QAR then CAViaR) simplifies implementation. Multiple quantile levels require repeated estimation. The adaptive ES approximation with iterative N selection adds computational overhead. R code with C++ functions is provided, reducing implementation burden. Key challenges include: (1) setting initial quantile values, (2) enforcing parameter constraints, (3) choosing the scale-defining quantile set P, and (4) the adaptive ES computation loop.

Data and code are publicly available at https://github.com/richardluger/QbSD. Code is written in R with C++ functions for computational efficiency. The paper provides detailed model specifications, estimation procedures, simulation designs, and empirical application parameters. Rolling window methodology is clearly described.