Abstract:We study the allocation of synthetic portfolios under hierarchical nested, one-factor, and diagonal structures of the population covariance matrix in a high-dimensional scenario. The noise reduction approaches for the sample realizations are based on random matrices, free probability, deterministic equivalents, and their combination with a data science hierarchical method known as two-step covariance estimators. The financial performance metrics from the simulations are compared with empirical data from companies comprising the S&P 500 index using a moving window and walk-forward analysis. The portfolio allocation strategies analyzed include the minimum variance portfolio (both with and without short-selling constraints) and the hierarchical risk parity approach. Our proposed hierarchical nested covariance model shows signatures of complex system interactions. The empirical financial data reproduces stylized portfolio facts observed in the complex and one-factor covariance models. The two-step estimators proposed here improve several financial metrics under the analyzed investment strategies. The results pave the way for new risk management and diversification approaches when the number of assets is of the same order as the number of transaction days in the investment portfolio.

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is how to effectively estimate the covariance matrix of portfolios with complex structures in high - dimensional scenarios and evaluate the performance of these estimators in different investment strategies. Specifically, the author focuses on: 1. **High - dimensional covariance matrix estimation**: When the number of assets \(p\) is on the same order of magnitude as the number of trading days \(n\), traditional covariance matrix estimation methods fail. Therefore, new estimators need to be developed to handle this high - dimensional scenario. 2. **Modeling of complex financial markets**: Financial markets exhibit complex interactions and dependencies, and these characteristics can be quantified through the covariance matrix. The author proposes a fully nested covariance model, aiming to capture these complexities. 3. **Optimization of investment strategies**: Through simulation and empirical data, evaluate the performance of different covariance estimators in strategies such as the minimum - variance portfolio (MVP) and hierarchical risk parity (HRP) to determine which estimator can better eliminate noise and improve the performance of the portfolio. ### Main research questions - **Can high - dimensional noise be effectively eliminated?**: Can the proposed estimators reveal the complex structures in the financial market? - **The impact of the denoising process on financial indicators**: Can the denoising process improve the performance of financial indicators such as HHI, L, and RDI? - **The sensitivity of different investment strategies to high - dimensional noise**: Which investment strategies are less sensitive to high - dimensional noise? ### Research methods - **Covariance matrix estimators**: Include multiple estimators based on random matrix theory (RMT), free probability, deterministic equivalence, and two - step estimation methods. - **Investment strategies**: Analyzed the minimum - variance portfolio (MVP), the minimum - variance portfolio without short - selling (MVP+), and the hierarchical risk parity (HRP) strategies. - **Performance evaluation**: Evaluate the performance of various estimators and investment strategies through simulation and empirical data (such as the daily closing prices of S&P 500 index companies). ### Conclusions The research shows that: - The proposed fully nested covariance model and the two - step estimator based on deterministic equivalence perform excellently on multiple financial indicators. - The HRP strategy is relatively insensitive to high - dimensional noise and shows good robustness. - Different estimators perform differently under different models, but overall they can effectively reduce noise and improve the risk management of portfolios. Through these studies, the author provides new ideas and tools for risk management in high - dimensional scenarios.

High-dimensional covariance matrix estimators on simulated portfolios with complex structures

A Dynamic Structure for High Dimensional Covariance Matrices and its Application in Portfolio Allocation

A Regularized High-Dimensional Positive Definite Covariance Estimator with High-Frequency Data

Large-Dimensional Positive Definite Covariance Estimation for High Frequency Data via Low-rank and Sparse Matrix Decomposition

Random matrix theory and nested clustered portfolios on Mexican markets

Element Aggregation for Estimation of High-Dimensional Covariance Matrices

Exponential Weighting and Random-Matrix-Theory-Based Filtering of Financial Covariance Matrices for Portfolio Optimization

Hierarchical DCC-HEAVY Model for High-Dimensional Covariance Matrices

Two-step estimators of high dimensional correlation matrices

Covariance Matrix Analysis for Optimal Portfolio Selection

High Dimensional Covariance Matrix Estimation Using Multi-Factor Models from Incomplete Information

Estimation of Large Financial Covariances: A Cross-Validation Approach

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

High dimensional minimum variance portfolio estimation under statistical factor models

Block-diagonal idiosyncratic covariance estimation in high-dimensional factor models for financial time series

High-Dimensional Volatility Matrix Estimation with Cross-Sectional Dependent and Heavy-Tailed Microstructural Noise

On the Modeling and Simulation of Portfolio Allocation Schemes: an Approach based on Network Community Detection

Estimation of Out-of-Sample Sharpe Ratio for High Dimensional Portfolio Optimization

High-Dimensional Conditional Covariance Matrices Estimation Using a Factor-GARCH Model

High-Dimensional Dynamic Panel with Correlated Random Effects: A Semiparametric Hierarchical Empirical Bayes Approach