In network time series, with errors correlated through the network at every time step but also correlated across time, estimating contemporaneous autocorrelations and those over time is a substantial challenge. Yet, for applications, understanding scenarios which guarantee consistent parameter estimation and scenarios in which some of these parameters cannot be consistently estimated, is needed to reach valid conclusions from data.

Attacking this problem requires both theoretical expertise and a range of suitable data sets for performance assessment. The project includes interactions with Financial NetworksAnalytics, with a particular emphasis on models for networks between banks in which correlations arise for example through inter-bank lending