Neural Network for CVA: Learning Future Values

2018/11/21 by She, Jian-Huang, Grecu, Dan
#Computational Finance (q-fin.CP) #FOS: Economics and business #Pricing of Securities (q-fin.PR)

paper · doi:10.48550/arxiv.1811.08726

Abstract

A new challenge to quantitative finance after the recent financial crisis is the study of credit valuation adjustment (CVA), which requires modeling of the future values of a portfolio. In this paper, following recent work in [Weinan E(2017), Han(2017)], we apply deep learning to attack this problem. The future values are parameterized by neural networks, and the parameters are then determined through optimization. Two concrete products are studied: Bermudan swaption and Mark-to-Market cross-currency swap. We obtain their expected positive/negative exposures, and further study the resulting functional form of future values. Such an approach represents a new framework for modeling XVA, and it also sheds new lights on other methods like American Monte Carlo.

Related

• Deep learning for CVA computations of large portfolios of financial\n derivatives
• Neural Networks for Portfolio-Level Risk Management: Portfolio Compression, Static Hedging, Counterparty Credit Risk Exposures and Impact on Capital Requirement
• Asset Pricing and Deep Learning
• Optimizing Neural Networks for Bermudan Option Pricing: Convergence Acceleration, Future Exposure Evaluation and Interpolation in Counterparty Credit Risk
• Deep Joint Learning valuation of Bermudan Swaptions