Informational Properties of Trading Networks

By Kimmo Soramäki | Published: November 3, 2009

There is a new interesting paper “On the Informational Properties of Trading Networks” by Adamic, Brunetti Harris and Kirilenko using transaction level data for all regular transactions in September 2008 E-mini S&P 500 futures contracts.

They construct networks from executed trades and then look at the structure of these networks and relate them to returns, volatility, volume, and duration. Interestingly they find that assortativity, low clustering coefficient and connectedness are positively related to returns and volume and negatively related to duration and volatility. In contrast, topologies with centralization and assortativity close to zero, high transitivity and high connectedness are associated with average returns and volatility, and positively related to volume and duration.

They construct networks from executed trades and then then look at the structure of these networks and related them to returns, volatility, volume, and duration of the trades.
. Interestingly they find that assortativity, low clustering coefficient and connectedness–are positively related to returns and volume and negatively related to duration and volatility. In contrast,
topologies with centralization and assortativity close to zero, high transitivity and high connectedness–are associated with average returns and volatility, and positively related to volume and duration.


Share|

This entry was posted in Research. Bookmark the permalink. Comments are closed, but you can leave a trackback: Trackback URL.