Probability of informed trading: an empirical application to the euro overnight market rate.

This paper presents a microstructure model for the unsecured overnight euro money market, similar to that developed for stock markets by Easley and O'Hara (1992). More specifically, this paper studies the role of heterogeneity in the population of banks participating on this market, and the influence of the institutional framework and market organizational aspects of the overnight deposit market. A first empirical assessment of the functioning of this market is based on the probability of informed trade which measures the ability of traders (banks) to interpret signals on the expected evolution of the overnight rate. This indicator is estimated on real-time data publicly available to market participants. Results show that between 2000 and 2004 a heterogeneous learning process of market mechanisms within participants could be observed. From 2005 onwards, however, heterogeneity in the learning process sharply decreased. Moreover, the empirical evidence show that the March 2004 changes in Eurosystem's operational framework have modified the informational patterns of order flow in the euro area money market: informed trades became even more predominant between the last main refinancing operation and the end of the reserves maintenance period than they were before March 2004.