Dynamic modelling of large-dimensional covariance matrices
Modelling and forecasting the covariance of financial return series has always been a challenge due to the so-called ‘curse of dimensionality’. This paper proposes a methodology that is applicable in large-dimensional cases and is based on a time series of realized covariance matrices. Some solutions are also presented to the problem of non-positive definite forecasts. This methodology is then compared to some traditional models on the basis of its forecasting performance employing Diebold—Mariano tests. We show that our approach is better suited to capture the dynamic features of volatilities and covolatilities compared to the sample covariance based models.
KeywordsSample Covariance Forecast Performance GARCH Model Exponentially Weight Move Average Sample Covariance Matrix
Unable to display preview. Download preview PDF.
- Aït-Sahalia Y, Mykland PA, Zhang L 2005 How often to sample a continuous-time process in the presence of market microstructure noise. Rev Financ Stud 18(2):351-416Google Scholar
- Andersen T, Bollerslev T, Christoffersen PF, Diebold FX (2006) Volatility forecasting. In: Elliott G, Granger C, Timmermann A (eds) Handbook of economic forecasting, 1st edition, chap 15, ElsevierGoogle Scholar
- Bandi FM, Russell JR (2005) Microstructure noise, realized volatility, and optimal sampling. Working paper, Graduate School of Business, The University of ChicagoGoogle Scholar
- Gourieroux C, Jasiak J, Sufana R (2004) The wishart autoregressive process of multivariate stochastic volatility. Working Paper, University of TorontoGoogle Scholar
- Oomen RCA 2005 Properties of bias-corrected realized variance under alternative sampling schemes. J Financ Econ 3:555-577Google Scholar
- Voev V, Lunde A 2007 Integrated covariance estimation using high-frequency data in the presence of noise. J Financ Econ 5:68-104Google Scholar