![[2402.06635] Large and Deep Factor Models](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2402.06635] Large and Deep Factor Models
Summary
The paper presents a deep neural network model for constructing stochastic discount factors in finance, highlighting a new component called the Portfolio Tangent Kernel that enhances pricing performance using U.S. equity data.
Why It Matters
This research is significant as it introduces a novel approach to factor modeling in finance, leveraging deep learning techniques. The findings can improve asset pricing strategies, which is crucial for investors and financial analysts aiming for better predictive accuracy in market behavior.
Key Takeaways
- Introduces a deep neural network model for stochastic discount factors.
- Highlights the Portfolio Tangent Kernel as a key component for pricing.
- Demonstrates significant performance gains using U.S. equity data.
- Discusses the impact of spectral complexity on pricing performance.
- Provides insights into the evolution of factor models since the early 2000s.
Quantitative Finance > Statistical Finance arXiv:2402.06635 (q-fin) [Submitted on 20 Jan 2024 (v1), last revised 13 Feb 2026 (this version, v2)] Title:Large and Deep Factor Models Authors:Bryan Kelly, Boris Kuznetsov, Semyon Malamud, Teng Andrea Xu, Yuan Zhang View a PDF of the paper titled Large and Deep Factor Models, by Bryan Kelly and 4 other authors View PDF HTML (experimental) Abstract:We show that a deep neural network (DNN) trained to construct a stochastic discount factor (SDF) admits a sharp additive decomposition that separates nonlinear characteristic discovery from the pricing rule that aggregates them. The economically relevant component of this decomposition is governed by a new object, the Portfolio Tangent Kernel (PTK), which captures the features learned by the network and induces an explicit linear factor pricing representation for the SDF. In population, the PTK-implied SDF converges to a ridge-regularized version of the true SDF, with the effective strength of regularization determined by the spectral complexity of the PTK. Using U.S. equity data, we show that the PTK representation delivers large and statistically significant performance gains, while its spectral complexity has risen sharply-by roughly a factor of six since the early 2000s-imposing increasingly tight limits on finite-sample pricing performance. Subjects: Statistical Finance (q-fin.ST); Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG) Cite as: arXiv:240...
