EUVIMEDEUVIMED entdecken
Crossref

Comparative Analysis of Pseudo-Random and Low-Discrepancy Number Generators: Implications for Normality and Value-at-Risk Estimation in Investment Portfolios

Sergei Kucherenko

Wilmott · 2026-05 · 2026 (143)

DOI: 10.54946/wilm.12236 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Wilmott, 2026-05. Verfasst von Sergei Kucherenko.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This study examines the impact of sampling schemes on Monte Carlo=based Value-at-Risk (VaR) estimation at both the risk factor and portfolio levels. The authors compare pseudo-random number generators with low-discrepancy sequences, including Sobol and Halton constructions. At the uncorrelated risk factor level, pseudo-random generators exhibit higher variability and less stable VaR estimates than low-discrepancy sequences. When correlation is introduced, the differences between the generators in how closely the theoretical VaR is achieved are substantially reduced. Moreover, portfolio aggregation further diminishes these differences, while the accuracy with which the theoretical VaR is represented also decreases across all generators. Optimized Sobol sequences (Sobol OPT and Sobol EN) demonstrate superior preservation of correlation structures, particularly in higher-dimensional settings. Overall, the results highlight the role of advanced random number generation techniques in enhancing the efficiency and reliability of VaR estimation in portfolio risk management.

Zitieren

APA

Kucherenko, S. (2026). Comparative Analysis of Pseudo-Random and Low-Discrepancy Number Generators: Implications for Normality and Value-at-Risk Estimation in Investment Portfolios. Wilmott, 2026(143). https://doi.org/10.54946/wilm.12236

Vancouver

Kucherenko S. Comparative Analysis of Pseudo-Random and Low-Discrepancy Number Generators: Implications for Normality and Value-at-Risk Estimation in Investment Portfolios. Wilmott. 2026;2026(143). doi:10.54946/wilm.12236

BibTeX
@article{Kucherenko2026,
  author = {Kucherenko, Sergei},
  title = {Comparative Analysis of Pseudo-Random and Low-Discrepancy Number Generators: Implications for Normality and Value-at-Risk Estimation in Investment Portfolios},
  journal = {Wilmott},
  year = {2026},
  volume = {2026},
  number = {143},
  doi = {10.54946/wilm.12236}
}
Kostenlos starten →