Accurate modeling and prediction of the financial asset covariance matrix plays an important role in building an effective portfolio. In this paper, a predictable matrix value factor model based on Cholesky decomposition and vector autoregressive method is adopted. This method significantly reduces the number of parameters to be estimated, effectively avoiding the accumulation of estimation errors and the dimensionality of low-dimensional matrices. Empirical analysis indicates that the dynamic investment portfolio can be obtained by constructing the investment strategy with minimum variance. The model can follow the market dynamics accurately and shows satisfactory forecasting ability, and the numerical experiments in the high-performance computing environment of supercomputer show that this model has good validity and scalability.
[1]
M. Damodaran,et al.
Parallel Three Dimensional Direct Simulation Monte Carlo for Simulating Micro Flows
,
2009
.
[2]
Marcelo C. Medeiros,et al.
Modeling and Forecasting Large Realized Covariance Matrices and Portfolio Choice
,
2017
.
[3]
Francis X. Diebold,et al.
Modeling and Forecasting Realized Volatility
,
2001
.
[4]
N. Shephard,et al.
Econometric Analysis of Vast Covariance Matrices Using Composite Realized Kernels and Their Application to Portfolio Choice
,
2016
.
[5]
Rong Chen,et al.
Factor models for matrix-valued high-dimensional time series
,
2016,
Journal of Econometrics.