Abstract:Most factor modelling research in vector or matrix-valued time series assume all factors are pervasive/strong and leave weaker factors and their corresponding series to the noise. Weaker factors can in fact be important to a group of observed variables, for instance a sector factor in a large portfolio of stocks may only affect particular sectors, but can be important both in interpretations and predictions for those stocks. While more recent factor modelling researches do consider ``local'' factors which are weak factors with sparse corresponding factor loadings, there are real data examples in the literature where factors are weak because of weak influence on most/all observed variables, so that the corresponding factor loadings are not sparse (non-local). As a first in the literature, we propose estimators of factor strengths for both local and non-local weak factors, and prove their consistency with rates of convergence spelt out for both vector and matrix-valued time series factor models. Factor strength has an important indication in what estimation procedure of factor models to follow, as well as the estimation accuracy of various estimators (Chen and Lam, 2024). Simulation results show that our estimators have good performance in recovering the true factor strengths, and an analysis on the NYC taxi traffic data indicates the existence of weak factors in the data which may not be localized.

What problem does this paper attempt to address?

### Problems Addressed by the Paper The paper primarily addresses the issue of estimating the strength of weak factors in vector and matrix time series factor models. Specifically: 1. **Importance of Weak Factors**: - Most factor modeling research assumes that all factors are strong (pervasive), classifying weaker factors and their corresponding variables as noise. However, weak factors can still be very important for specific groups of observed variables in certain situations. For example, in a large stock portfolio, a factor related to a particular industry may only affect stocks within that industry but remains important for explaining and predicting those stocks. 2. **Limitations of Existing Methods**: - Recent studies have started to consider "local" weak factors (i.e., weak factors with sparse factor loadings), but there are cases in real data where non-local weak factors exist. These factors have a weak influence on most or all observed variables, but their corresponding factor loadings are not sparse. 3. **Proposed Method**: - This paper is the first to propose a method for estimating the strength of both local and non-local weak factors and demonstrates its convergence in vector and matrix time series factor models. Additionally, simulation experiments validate the proposed method's good performance in recovering true factor strength, and an analysis of NYC taxi flow data shows the presence of non-local weak factors in the data. 4. **Theoretical Contribution**: - A new method is proposed to estimate factor strength in vector and matrix time series factor models without relying on the sparsity assumption of the factor loading matrix. By directly extracting factor strength using covariance information and the estimated factor loading matrix, this method is applicable for general cases of factor strength estimation, especially when the sparsity assumption is not relied upon. By addressing the above issues, the paper provides an important theoretical foundation and practical tools for factor strength estimation, helping to improve the accuracy and interpretability of factor models.

Factor Strength Estimation in Vector and Matrix Time Series Factor Models

Time-Varying Matrix Factor Models

Rank and Factor Loadings Estimation in Time Series Tensor Factor Model by Pre-averaging

Factor Models for Matrix-Valued High-Dimensional Time Series

Rank Determination in Tensor Factor Model

On Time-Varying Factor Models: Estimation and Testing

Knowing Factors or Factor Loadings, or Neither? Evaluating Estimators of Large Covariance Matrices with Noisy and Asynchronous Data

Modeling High-Dimensional Time Series: A Factor Model With Dynamically Dependent Factors and Diverging Eigenvalues

Approximate Factor Models with Strongly Correlated Idiosyncratic Errors

Factor Models for Portfolio Selection in Large Dimensions: The Good, the Better and the Ugly

Vector or Matrix Factor Model? A Strong Rule Helps!

Matrix-valued Factor Model with Time-varying Main Effects

Constrained Factor Models for High-Dimensional Matrix-Variate Time Series

When can weak latent factors be statistically inferred?

Inference in Sparsity-Induced Weak Factor Models

Factor Models for High-Dimensional Tensor Time Series

Improving Estimation of Portfolio Risk Using New Statistical Factors

The Dynamic, the Static, and the Weak factor models and the analysis of high-dimensional time series

High Dimensional Factor Analysis with Weak Factors

Estimation of Sparsity-Induced Weak Factor Models