Kernel Autocorrelogram for Time Deformed Processes

The purpose of the paper is to propose an autocorrelogram estimation procedure for irregularly spaced data which are modelled as subordinated continuous time series processes. Such processes, also called time deformed stochastic processes, have been discussed in a variety of contexts. Before entertaining the possibility of modelling such time series one is interested in examining simple diagnostics and data summaries. With continuous time processes this is a challenging task which can be accomplished via kernel estimation. This paper develops the conceptual framework, the estimation procedure and its asymptotic properties. An illustrative empirical example is also provided.
[ - ]
[ + ]