Risk factor aggregation and stress testing

Paper: arXiv 2310.04511

Authors: Natalie Packham

Abstract

Stress testing refers to the application of adverse financial or macroeconomic scenarios to a portfolio. For this purpose, financial or macroeconomic risk factors are linked with asset returns, typically via a factor model. We expand the range of risk factors by adapting dimension-reduction techniques from unsupervised learning, namely PCA and autoencoders. This results in aggregated risk factors, encompassing a global factor, factors representing broad geographical regions, and factors specific to cyclical and defensive industries. As the adapted PCA and autoencoders provide an interpretation of the latent factors, this methodology is also valuable in other areas where dimension-reduction and explainability are crucial.

Complexity vs Empirical Score

• Math Complexity: 5.5/10
• Empirical Rigor: 6.5/10
• Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper presents a solid application of unsupervised learning techniques (PCA and autoencoders) to risk factor aggregation and stress testing in portfolios. It balances linear and non-linear dimension reduction with practical financial explainability, supported by reasonable empirical modeling.

Research Flowchart

  flowchart TD
    A[Research Goal: Expand Stress Testing Risk Factors via Unsupervised Learning] --> B[Input Data: Financial & Macroeconomic Risk Factors / Asset Returns]
    B --> C[Dimension Reduction: Adapted PCA & Autoencoders]
    C --> D[Factor Modeling: Link Latent Factors with Asset Returns]
    D --> E[Outcome: Interpretable Aggregated Factors<br>Global, Regional, Cyclical & Defensive]
    E --> F[Application: Explainable Stress Testing & Risk Aggregation]