Nonparametric sign prediction of high-dimensional correlation matrix coefficients
arXiv:2001.11214 · doi:10.1209/0295-5075/133/48001
Abstract
We introduce a method to predict which correlation matrix coefficients are likely to change their signs in the future in the high-dimensional regime, i.e. when the number of features is larger than the number of samples per feature. The stability of correlation signs, two-by-two relationships, is found to depend on three-by-three relationships inspired by Heider social cohesion theory in this regime. We apply our method to US and Hong Kong equities historical data to illustrate how the structure of correlation matrices influences the stability of the sign of its coefficients.
References in corpus (19)
- Fast unfolding of communities in large networks
- Hierarchical Structure in Financial Markets
- Noise Dressing of Financial Correlation Matrices
- Intensity and coherence of motifs in weighted complex networks
- A Random Matrix Approach to Cross-Correlations in Financial Data
- A tool for filtering information in complex systems
- Clustering and information in correlation based financial networks
- Social Balance on Networks: The Dynamics of Friendship and Enmity
- Triadic closure as a basic generating mechanism of communities in complex networks
- Statistically validated networks in bipartite complex systems
- Pathways towards instability in financial networks
- Cleaning large correlation matrices: tools from random matrix theory
- Complex Networks on Hyperbolic Surfaces
- Emergence of time-horizon invariant correlation structure in financial returns by subtraction of the market mode
- Evolution of correlation structure of industrial indices of US equity markets
- Mesoscopic Community Structure of Financial Markets Revealed by Price and Sign Fluctuations
- Core of communities in bipartite networks
- Covariance matrix filtering with bootstrapped hierarchies
- Binary versus non-binary information in real time series: empirical results and maximum-entropy matrix models