cCorrGAN: Conditional Correlation GAN for Learning Empirical Conditional Distributions in the Elliptope

Abstract

We propose a methodology to approximate conditional distributions in the elliptope of correlation matrices based on conditional generative adversarial networks. We illustrate the methodology with an application from quantitative finance: Monte Carlo simulations of correlated returns to compare risk-based portfolio construction methods. Finally, we discuss about current limitations and advocate for further exploration of the elliptope geometry to improve results.

arXiv commentInternational Conference on Geometric Science of Information

References in corpus (4)

• Quant GANs: Deep Generation of Financial Time Series
• Derivatives of Multilinear Functions of Matrices
• CorrGAN: Sampling Realistic Financial Correlation Matrices Using Generative Adversarial Networks
• Clustering in Hilbert simplex geometry