Researchers propose adapted methodology to assess money market microstructure

A working paper published by the Bank of Italy proposes an adaptation of the standard methodology used to identify bilateral exposures in the interbank loan market.

The adapted model is set out in Estimating the money market microstructure with negative and zero interest rates, by Edoardo Rainone and Francesco Vacirca. They note an algorithm devised by Craig Furfine in a 1999 paper has been widely used in subsequent research efforts to identify such exposures.

All the main interpretations of Furfine's algorithm, they say, "implicitly assume that key interest rates are strictly positive". Furthermore, they argue, "the standard version of the Furfine algorithm is not able to detect loans traded at a zero rate with a good degree of reliability".

They set out an economic methodology that uses regularities observed in decentralised markets. This information, they say, allows them to distinguish loans from "false loans", or pairs of payments not related to lending activity.

They then apply this adapted methodology to data from Target-2. Data from this test shows, the authors argue, that effects of negative interest rates on the interbank lending market is "quite different" to those predicted by the standard Furfine model.