Diffusion Factor Models: Generating High-Dimensional Returns with Factor Structure
Minshuo Chen, Renyuan Xu, Yumin Xu, and Ruixun Zhang · 2025
Read Diffusion Factor Models: Generating High-Dimensional Returns with Factor Structure by Minshuo Chen, Renyuan Xu, Yumin Xu, and Ruixun Zhang (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-06566.txt.
Description
Minshuo Chen and co-authors build a new tool for faking market data in this 2025 study. Banks need fake scenes of stock returns to test risk and tune portfolios, but real data are scarce and cover thousands of stocks at once. Their method joins a modern image-style generator with an old idea from finance: a few hidden factors drive most price moves.
By building the factor shape into the network, the method learns from the small factor count, not the huge stock count. The authors prove error bounds that rest on that small count. Tests showed the tool recovers hidden market groups well from little data. Portfolios built from its fake scenes also held up in real market checks.
Catalog record
| CREATOR | Minshuo Chen, Renyuan Xu, Yumin Xu, and Ruixun Zhang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 242 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-06566