- Dynamic Clustering of Exchange Rates
- Communities in Networks
- Community structure in time-dependent, multiscale, and multiplex networks
- Community Structure in Congressional Co-Sponsorship Networks

If one was interested in the time evolution of community structure in inter-bank networks with an eye towards creating indicators of systemic risk...then there are a couple of issues that would need to be addressed:

- Develop a robust network data set! This type of analysis is incredibly data intensive. You need enough of a network data so that you can take cross-section slices at various time intervals (quarterly, monthly, weekly, etc). Ideally you would need to take daily slices in order to develop systemic risk indicators that would be useful...
- Develop a MEANINGFUL algorithm to identify community structure in bank networks. There are a number of algorithms that have been developed to id community structure. However, as discussed in Matthew Jackson's book on social networks, it is often difficult to identify exactly these algorithms are capturing (i.e., do the identified communities have any economic justification? Or are they simply a statistical artifact of the data?)
- What network analysis techniques are must appropriate to use as measures of systemic risk?
- Last but not least, how do you get buy-in from clients that this type of analysis would be useful?

## No comments:

## Post a Comment