Merton portfolio optimization with a Wishart-process covariance (Dyson eigenvalue repulsion / RMT), solved via matrix Riccati and a Deep BSDE. 170 tests + CI.
-
Updated
Jun 27, 2026 - Python
Merton portfolio optimization with a Wishart-process covariance (Dyson eigenvalue repulsion / RMT), solved via matrix Riccati and a Deep BSDE. 170 tests + CI.
Clean-room PyTorch replication of Han-Jentzen-E (PNAS 2018) Deep BSDE, applied to the multi-asset Black-Scholes PDE. 100-D basket call priced to 0.22% vs 10^7-path MC.
Add a description, image, and links to the deep-bsde topic page so that developers can more easily learn about it.
To associate your repository with the deep-bsde topic, visit your repo's landing page and select "manage topics."