Tuesday June 4 2013
Scuola Normale Superiore
Department of Statistics & Applied Probability – National University of Singapore
We develop a synchronizing technique for irregularly spaced and asynchronous high frequency data. The technique learns from the dependence structure of raw data and iteratively recovers the unobserved values of the synchronous series at high sampling frequency.
The numerical results illustrate the performance of the proposed technique and compared to the conventional techniques — Previous Tick technique and Refresh Time technique. The proposed technique provides good performance in terms of accuracy and feature.
Moreover, a realized covariance estimator is constructed by incorporating the synchronized technique. We compare the feature of the estimator with several alternative estimators.