Forecasting Large Realized Covariance Matrices: The Benefits of Factor Models and Shrinkage*.

Alves, Rafael P; Brito, Diego S de; Medeiros, Marcelo C; Ribeiro, Ruy M

We propose a model to forecast large realized covariance matrices of returns, applying it to the constituents of the S&P 500 daily. To address the curse of dimensionality, we decompose the return covariance matrix using standard firm-level factors (e.g. size, value, and profitability) and use sectoral restrictions in the residual covariance matrix. This restricted model is then estimated using vector heterogeneous autoregressive models with the least absolute shrinkage and selection operator. Our methodology improves forecasting precision relative to standard benchmarks and leads to better estimates of minimum variance portfolios.

COVARIANCE matrices; FORECASTING; AUTOREGRESSIVE models; STANDARD & Poor's 500 Index; FORECASTING methodology

Journal of Financial Econometrics, 2024, Vol 22, Issue 3, p696

1479-8409

Academic Journal

10.1093/jjfinec/nbad013