Hierarchical and Adaptive Methods for Efficient Risk Estimation

该项目旨在扩展现有的适应性方案，以有效地计算重要的风险度量，并将其应用于金融。特别是，我们的目标是将这些方法应用于信用价值调整(CVA)资本费用-该费用要求银行预留资金，以防止因一方拖欠银行款项而无法偿还所欠款项而造成的损失。CVA费用是巴塞尔III协议的关键组成部分，该协议旨在缓解在2008年金融危机中发挥关键作用的风险。CVA费用显式公式的性质和影响费用的各种变量使得标准计算方法不可行。银行经常求助于使用廉价、不准确的近似值。我们的目标是改进精确计算方法的复杂性，为这些廉价的近似提供一个有效、准确的替代方案。我们还计划将这项研究扩大到其他(所谓的XVA)金融风险衡量标准中。我们的目标也是推广相同的方法，以纳入各种具有计算困难结构的金融合同。

This project aims to expand upon existing adaptive schemes to efficiently compute important measures of risk with applications to finance. In particular, we aim to apply these methods to the Credit Value Adjustment (CVA) Capital Charge - which requires banks to reserve money to protect against losses resulting from a party who is owe money to the bank defaulting and not repaying money owed. The CVA charge is a key component in the Basel III accord which aims to mitigate the risks which played a key role in the 2008 financial crash. The nature of the explicit formula for the CVA charge and the variety of variables influencing the charge render standard computational methods infeasible. Banks often resort to using cheap, inaccurate approximations instead. It is our goal to improve upon the complexity of accurate computational methods to provide an efficient, accurate alternative to these cheap approximations. We also plan to expand the research for use in other (so-called xVA) financial risk-measures. It is also our goal to generalize the same approach to incorporate various financial contracts with computationally difficult structure.