Econometric analysis of vast covariance matrices using composite realized kernels∗

We propose a composite realized kernel to estimate the ex-post covariation of asset prices. Composite realized kernels are a data efficient method where the covariance estimate is composed of univariate realized kernels to estimate variances and bivariate realized kernels to estimate correlations. We analyze the merits of our composite realized kernels in an ultra high dimensional environment, making economic decisions every day solely based on the previous day’s data. The first application is a minimum variance portfolio exercise and this is followed by an investigation of portfolio tracking. The data set is tickby-tick data comprising 473 US equities over the sample period 2006-2009. We show that our estimator is able to deliver a significantly lower portfolio variance than its competitors.